Skip to main content

Full text of "Journal of the Society of Motion Picture Engineers"

See other formats

From the collection of the 

7 n 

o Prelinger h 
v Jjibrary 

San Francisco, California 

Journal of the Society of 

Motion Picture and Television Engineers 


Simplification of Motion Picture Processing Methods 

C. E. IVES and C. J. KUNZ 3 

A 16-Mm Rapid Film Processor 

A Method of Measuring Electrification of Motion Picture Film Applied to 

Cleaning Operations H. W. CLEVELAND 37 

Variable- Area Sound Track Requirements for Reduction Printing Onto Koda- 

chrome ROBERT V. McKiE 45 

The Pressurized Ballistics Range at the Naval Ordnance Laboratory 

An Experimental Electronic Background Television Projection System. . . . 

Effects of Incorrect Color Temperature on Motion Picture Production .... 


The Stroboscope as a Light Source for Motion Pictures 


Study of Sealed Beam Lamps for Motion Picture Set Lighting 


Color Committee Report HERMAN H. DUERR 113 


Scanning-Beam Uniformity Test Film for 16-Mm Sound Reproducers 
(Laboratory Type), Z22.80-1950; Scanning-Beam Test Film for 16- 
Mm Sound Reproducers (Service Type), Z22.81-1950; Sound Trans- 
mission of Theater Projection Screens, Z22.82. 

68th Semiannual Convention 121 

High-Speed Photography Question Box 122 

Engineering Committees Activities 123 

Society Announcements 123 

New Members 124 

LETTER TO THE EDITOR By Joseph H. Spray 125 

BOOK REVIEW: Handbook of Basic Motion-Picture Techniques by Emil E. 

Brodbeck Reviewed by James W. Moore 126 

New Products 127 

Employment Service 128 


Chairman Editor Chairman 

Board of Editors Papers Committee 

Subscription to nonmembers, $12.50 per annum; to members, $6.25 per annum, included in 
their annual membership dues; single copies, $1.50. Order from the Society's General Office. 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and 
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture and Tele- 
vision Engineers, Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General 
and Editorial Office, 342 Madison Ave., New York 17, N.Y. Entered as second-class matter 
January 15, 1930, at the Post Office at Easton, Pa., under the Act of March 3, 1879. 
Copyright, 1950, by the Society of Motion Picture and Television Engineers, Inc. Permission 
to republish material from the JOURNAL must be obtained in writing from the General Office of 
the Society. Copyright under International Copyright Convention and Pan-American Con- 
vention. The Society is not responsible for statements of authors or contributors. 

Society of 

Motion Picture and Television Engineers 

342 MADISON AVENUE, NEW YORK 17, N.Y., TEL. Mu 2-2185 


Earl I. Sponable 
460 W. 54 St. 
New York 19, N.Y. 

Peter Mole 

941 N. Sycamore Ave. 
Hollywood 38, Calif. 


Loren L. Ryder 
5451 Marathon St. 
Hollywood 38, Calif. 

Clyde R. Keith 
120 Broadway 
New York 5, N.Y. 

Robert M. Corbin 
343 State St. 
Rochester 4, N.Y. 

William C. Kunzmann 
Box 6087 
Cleveland 1, Ohio 

Fred T. Bowditch 
Box 6087 
Cleveland 1, Ohio 


Frank E. Cahill, Jr. 
321 W. 44 St. 
New York 18, N.Y. 

Ralph B. Austrian 
25 W. 54 St. 
New York 19, N.Y. 


Herbert Barnett 
Manville Lane 
Pleasantville, N.Y. 

Kenneth F. Morgan 
6601 Romaine St. 
Los Angeles 38, Calif. 

Norwood L. Simmons 
6706 Santa Monica Blvd. 
Hollywood 38, Calif. 


Charles R. Daily 
113 N. Laurel Ave. 
Los Angeles 36, Calif. 

John P. Livadary 
4034 Cromwell Ave. 
Los Angeles, Calif. 

William B. Lodge 
485 Madison Ave. 
New York 22, N.Y. 


Lorin D. Grignon 
1427 Warnall Ave. 
Los Angeles 24, Calif. 

Paul J. Larsen 
6906 S. Bradley Blvd. 
Bradley Hills 
Bethesda, Md. 

William H. Rivers 
342 Madison Ave. 
New York 17, N.Y. 


F. E. Carlson 
Nela Park 
Cleveland 12, Ohio 

George W. Colburn 
164 N. Wacker Dr. 
Chicago 6, 111. 

Edward Schmidt 

304 E. 44 St. 

New York 17, N.Y. 

Malcolm G. Townsley 
7100 McCormick Rd. 
Chicago 45, 111. 

Edward S. Seeley 
161 Sixth Ave. 
New York 13, N.Y. 

R. T. Van Niman 
4501 Washington St. 
Chicago 24, 111. 

Simplification of Motion Picture 
Processing Methods 


SUMMARY: The chemical bath formulas and treating methods used in 
present-day continuous motion picture processing machines were adopted 
without essential modification from the earlier manually operated rack-and- 
tank process, to which the long times of treatment were well suited. In 
continuous processing at high running speed, these long times of treatment 
require the use of large-size machines of considerable complexity which are 
costly to build and difficult to operate and maintain. Recent work on rapid 
processing methods has shown that, with highly active baths and spray ap- 
plication, the tunes of treatment can be reduced by a factor of 25 to 50, so 
that equipment can be made smaller and simpler as well as easier to operate 
and maintain. 

With such types of film as can be strongly hardened in manufacture, 
elevated temperatures are used to accelerate the reactions further and to 
simplify temperature control without refrigeration. In this case, processing 
is complete in a minute or less. Even with films which are not hardened 
to such a degree in manufacture, the total time for processing can usually 
be reduced to a few minutes by making use of active baths applied by spray- 
ing and impingement warm-air drying. The latter films sometimes are 
hardened in a preliminary bath to gam time by the use of vigorous baths 
and elevated temperatures if the process comprises a number of successive 
bathing operations. 

The design of equipment to suit the needs of these processing methods is 
described with reference to the conditions which are met in television work, 
in the motion picture laboratory and in the field. 

AMONG ALL the types of photography, motion picture work stands 
out as that in which the degree of mechanization and the com- 
pleteness of technical control during the processing operations are 
greatest. The present high level of quality and uniformity of re- 
sults bear evidence of the effectiveness of the effort which has been 
made to improve materials and equipment over the years. While 
quality has undergone continuous improvement during the last three 
decades, the running speed of processing machines has increased by a 
factor of several times, with a corresponding gain in productive ca- 

The chemical processes around which the continuous machine is 
built, however, have not changed significantly from the day of the 
hand-manipulated rack-and-tank method for which they were de- 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 
Communication No. 1317 from the Kodak Research Laboratories. 



vised. It is probably fair to say that the chemical process has no1 
been adapted in any way to take advantage of the capabilities o: 
automatic machinery for rapid, precise operation, and that no recog- 
nizable trend in machine design has demanded any important modifi- 
cation of the process. Consequently, the present fast-running ma- 
chine, with its hundreds of transport rollers along a film path thou- 
sands of feet in length, is so large and complex that it requires mucl: 
skill for operation and maintenance and the use of a large amount &. 
associated equipment. Nevertheless, existing equipment of this 
type is meeting the requirements of the large laboratories, with re- 
spect to both the quality and quantity produced, and probably wil 
do so for some time to come. On the other hand, this type of equip- 
ment is too bulky and inflexible for use in certain special applications 
The present work is concerned primarily with the latter cases in 
which requirements are unusual, but the results obtained are sig- 
nificant in many respects for general processing practice as well. 

Although some simplification in the operation of processing equip- 
ment could be achieved by redesigning individual elements and intro- 
ducing more elaborate control instruments, much more might be ac- 
complished if the time of treatment in the various steps of the proc- 
ess could be shortened by a factor of ten times or more. Even in 
cases where a reduction of the time of processing is not in itself oi 
paramount importance, shortening of the film path and diminution 
in the volume of the baths would permit radical changes in design 
and in methods to provide more automatic operation. 

About twenty years ago it appeared that a photographic film 
intermediate step would be required in television for the sake of the 
additional sensitivity it offered in the pickup from original sub- 
jects and in some cases for light amplification and image storage at 
the receiving point. In response to these needs, considerable work 
was done on rapid, highly automatic processing methods, 1 " 3 some of 
which have found application in other types of photography 4 - 5 and 
appear to offer promise in present-day motion picture work. More 
recently, the requirements of military use have led to the develop- 
ment of stepwise processing methods 6 ' 7 in which the time of treat- 
ment was shortened by a factor of 25 to 50 times, compared to that of 
ordinary practice, by the use of highly active baths, elevated tem- 
peratures, forceful application of the baths and, to a limited extent, 
special photographic films. 

The strenuous treatment employed to obtain the most rapid proc- 
essing in some of the cases cited would cause serious softening of the 
emulsion gelatin unless it was hardened either in manufacturing or at 


a suitable stage in the processing. At the time when further study 
of applications of these methods to motion picture work was under- 
taken, the hardening of some of the lower-speed printing and sound- 
recording films had been increased sufficiently in manufacture and 
effective methods were available for hardening other films in proc- 
essing. Preliminary studies of washing and drying techniques had 
indicated that great acceleration of the process was possible by 
adoption of forceful jet impingement methods in place of the low ve- 
locities and slow renewal characteristic of current practice. 


Work on rapid processing published up to the present time has 
revealed little as to the uniformity and image quality attainable for 
motion picture use. Since temperatures would often be well above 
ambient and short times of treatment would prevail with the methods 
considered, automatic equipment with thermostatic control would be 
needed for the investigation. Also, in due course, practical equip- 
ment would be required for studies of the techniques for applying 
treating baths and drying air. 

Two major units of continuous processing equipment were there- 
fore built and used in this work, although supplementary tests were 
carried out on a variety of other equipment. The first machine was a 
highly compact, semiportable unit occupying about 2 cu ft of space, 
and the second was intended for operation at the rate of 90 fpm and 
was proportionately larger. Based on preliminary tests, the design 
in both cases provided for the complete processing of the highly 
hardened Eastman Fine Grain Release Positive Film in a minute or 

Semiportable 16-Mm Machine 

The first continuous machine used in the present work was de- 
signed to give about 5-sec immersion treatment in developer, rinse, 
fixing and washing tanks, respectively, at a running speed of 8 fpm. 
It was intended to be highly compact, easy to thread, and simple in 
construction. Figure 1 shows the machine with tanks removed, re- 
vealing the film in normal running position. Threading is accom- 
plished by drawing the film from the supply box at the lower left across 
the upper rollers standing between the several tank compartments 
into which loops are formed when the rack assembly carrying the 
lower rollers is slid down into position. Upon leaving the last tank 
at the right, the film passes through the squeegee rollers and then 
around the two large heated drive drums on which drying is effected 



Fig. 1. 16-Mm continuous rapid processing machine, with tank section re- 
moved to show film path. 



with the help of compressed air discharged against the emulsion sur- 
face through orifices in the arcuate distributing pipes. The dried 
film is wound up on the reel at the top left. When the rack frame is 
lowered to form the loops, the drive is automatically connected by 
radial engagement of a spur gear on the shaft of one of the interlinked 
drums. One ounce of processing bath is then run into each tank and 
brought to working temperature by a thermostatically controlled 
heating element immersed in the water jacket, which can be seen 
with the tank assembly in Fig. 2. When the working temperature is 

Fig. 2. Tank section for 16-mm rapid processing machine; heat-exchanger 
jacket at left, and four processing tanks with preheating coils at right. 

reached, as evidenced by a heater pilot lamp bull's-eye, wash water 
and compressed air are turned on and the drive is started. Developer 
and fixing-bath replenishers flow continuously under control of 
throttle valves from the constant-level chambers above at the back 
and reach the work tanks after passing through tempering coils 
(Fig. 2) in the water jacket. Just below the drying drum at the 
right (Fig. 1) is a transparent box enclosing an additional film loop 
and two water spray nozzles which can be used when thorough wash- 
ing is desired or by-passed if a minimum processing time is required. 


At a film speed of 8 fpm, a 4.8-sec time of treatment is provided in 
each bath, which is sufficient for normal processing at 125 F (degrees 
Fahrenheit) when Kodak D-8 Developer and the Kodak Rapid Liquid 
Fixer (with Hardener) are used with Eastman Fine Grain Release 
Positive Film, Type 7302. Except in the spray wash section, the 
baths are neither agitated nor circulated except as a result of film 

While the rate of reaction in the chemical baths was not affected 
significantly by the lack of agitation, development uniformity and 
tone reproduction were not always up to commercial standards for 
continuous-tone work. Hot-drum drying, while efficient from the 
viewpoint of heat transfer, was prone to cause mottle pattern be- 
cause of the practically unavoidable nonuniformity of contact. 

Nevertheless, this apparatus has served its purpose well in a great 
variety of experimental work involving a wide range of temperatures, 
chemical treating methods, and film types. The machine was suf- 
ficiently light-tight for daylight operation and could be loaded with- 
out the use of a magazine if the film had removable opaque backing. 
The high-pH developers were effective in removal of the backing ma- 
terial with the aid of a light frictioning in the bath. 

With the addition of a small air-compressing pump and a pressure 
tank for the water supply, this equipment required only a source of 
electric current for installation almost anywhere. Therefore, it has 
been sent out frequently for use in tests involving other less portable 
equipment, 8 for processing in an airplane, and for lecture-hall demon- 
strations, and has given regular service in laboratory work. 

A 90-fpm Machine 

In order to evaluate rapid processing for such applications as 
theater television and motion picture laboratory work where highest 
standards of quality and uniformity would have to be met, a faster 
machine with more effective means for bath agitation and for drying 
was needed. A 90-fpm speed was decided upon with intense spray 
application of chemical solutions and wash water to secure the rapid 
and uniform renewal at the film surface which would be requisite 
with the short times of treatment. 

Provision was made for times of treatment as short as 5 sec in each 
bath, with the possibility of increase to 10 or 15. The film was sup- 
ported on a cylindrical drum during the 5-sec drying treatment so 
that forcible air jets could be applied. Longer times would be at- 
tainable by reducing the running speed. Unless the film splicing and 
roll changing could be made fast and entirely automatic, the usual 




type of elevator and splicer sections would be disproportionately 
large in comparison with the rest of the machine. These elements 
were, therefore, omitted entirely, inasmuch as a machine with such a 
short processing time and film path could be stopped at the end of 
each roll, at least in experimental work. 

Arrangement of Parts 

In the front elevation of the machine in operating condition (Fig. 3) 
starting from the left are seen the supply roll, the developer cabinet, 


Fig. 3. 90-Fpm 35- and 16-mm rapid processing machine. 

a short section for rinsing, the fixing and washing cabinets, the pneu- 
matic squeegee, the drying drum, and at the top right, the film-drive 
roller and the windup. Figure 4 is a closer view of the bathing sec- 
tion with the doors opened as for threading to show the arrangement 
of the film loops and spray nozzles. These figures are essentially the 
same as those shown at the SMPE Chicago Convention in the spring 


of 1947, when the method was discussed in a preliminary way. This 
machine is about 6 ft long, 7 h high, and 2 ft deep. 

In the three larger cabinets, the film traverses the familiar flattened 
helical path over free-running rollers supported by the parallel upper 
and lower shafts on 42-in. vertical centers. The lower shaft is fixed 
in position and does not rotate, while the upper shaft is mechanically 
driven at a speed somewhat greater than that at which the rollers 
are turning with the moving film. With the upper shaft somewhat 
larger than the lower, this overdrive of a few percent largely neu- 
tralizes the frictional drag and relieves film tension. The film moves 
through the small rinse cabinet in a straight line so as to provide a 
time of about one-half second. 

Figure 5 shows the circulatory path of the developer and fixer. 
From the sump at the bottom of the rectangular tank at the upper 
left, the developer goes through the pump, a heater and a filter, past the 
thermostatic switch and thermometer to the spray nozzle system. In 
order to simplify piping while obtaining complete spray coverage of 
the film, the nozzles were located inside the film loops and directed 
upward and transversely to the film at an angle of about 30 to the 
vertical. It had been determined in advance that films which are 
as strongly hardened in manufacture as Eastman Fine Grain Release 
Positive Film, Type 5302 (or 7302), could safely be run with the emul- 
sion in contact with a reasonable number of smooth rollers, as long as 
the film surface was kept completely wet and proper principles were 
adhered to in design and maintenance. Stainless-steel construction 
and piping were used, where necessary, with commercially available 
spray nozzles of the same material. 

The pneumatic squeegee used was of a type employed on many 
conventional processing machines and consisted of a hollow box with 
roller-guarded slotted openings at opposite ends. 9 

While a considerable variety of drying schemes were of interest, 
the basic equipment on this machine consisted of distribution piping 
for the drying-air jets and a radiant-heating ribbon of Ni chrome sur- 
rounding the drum and concentric with it and at a distance of about 
one-half inch from the film. In some of the work, a cabinet dryer 
located above the drum was used instead. 

Applications of Spray-Type Machine 

Although the ability of this type of equipment to do good work had 
been demonstrated prior to the preliminary report presented in the 
spring of 1947, much remained to be learned both as regards fea- 
tures of the equipment design and as to the application of a rapid 




processing technique to films of widely varying properties. Because 
of the lack of automatic equipment for preliminary study on a test- 
tube scale and the need for information on processing-machine design, 
the experimental work was carried out mainly with the continuous 
machines under conditions of practical use and will, therefore, be 
discussed on the same basis. 

Processing Cobinet Pressure. Goug 

Level of Solution 
in Sump 


Filter Swifch 


Fig. 4. Spray processing chambers of 90- 
fpm machine, with doors opened to show 
position of nozzles. 

Fig. 5. Circulation system in 
90-fpm rapid processing machine. 


In the processing of Eastman Fine Grain Release Positive Film in 
theater television use, both compactness of equipment and extreme 
curtailment of the processing time were required. A study was 
made, therefore, of processing methods for use in the 90-fpm spray- 
type machine, with the object of arriving at a ^-min total time for 
processing, including drying. 


The 40-fold reduction in developing time from 3J^ min to 5 sec was 
achieved by the combined effect of a very active developer, such as 
Kodak SD-27 (formula below), and an elevation of temperature from 
70 to 120 F. 

Kodak Rapid Developer SD-27 

Water, about 90 F (32 C) ....................... 750.0 ml 

Kodak Elon Developing Agent .................. 5 . g 

Kodak Hydroquinone .......................... 45 . g 

Kodak Sodium Sulfite, desiccated ................ 90.0 g 

Kodak Sodium Hydroxide (Caustic Soda) ......... 40 . g 

Kodak Potassium Bromide ...................... 10 . g 

Kodak Antir-Fog No. 1 (Benzotriazole) ........... l.Og 

Water to make ................................ 1.01 

Some difficulty was experienced with nonuniformity of density 
until it was realized that complete wetting of the film must be at- 
tained in the first one-half second which, of course, constituted 10% 
of the total developing time. The use of a wetting agent, such as 
"Tergitol" Penetrant 08 (Manufactured by Carbide & Carbon 
Chemicals Div., Union Carbide & Carbon Corp., New York, N.Y.) 
at a concentration of 0.1% in the developer was helpful but insufficient 
except in combination with proper application of the bath at the 
start. The practice finally adopted consisted in momentary im- 
mersion in developer contained in a small vestibular trough at the 
cabinet entrance followed immediately by a strong spray blast. 
The wetting agent was retained for its effect in preventing a type of 
marking caused by small scattered airbells. Developer from the first 
nozzle maintained sufficient depth in the trough to insure wetting of 
the entering film and sealing of the opening against the entry of air. 
Additional pressure of developer at the nozzles above the minimum 
of 30 psi required to produce the full spray pattern had no effect on 
the quality of results. 


Adequate rinsing was obtained in the 9-in.-long straight pass 
through the next small cabinet, which was equipped with two spray 
nozzles directed upward toward the emulsion surface and one down- 
ward toward the film support. All three were supplied with tempered 
water. Any deficiency in rinsing tended to cause the typical yellow- 
ish fog which is formed when residues of a vigorous developer are 
present in film as it enters a fixing bath. Soft-rubber wringer rollers 
located at the entrance and exit openings of each cabinet are helpful 
in minimizing carry-over and leakage. 



Complete fixation of the Eastman Fine Grain Release Positive 
Film in 10 sec at 120 F, was attained by the use of an ammonium thio- 
sulfate bath, such as Kodak Rapid Liquid Fixer (with Hardener), at a 
dilution of one part of the commercially supplied concentrate to 
three parts of water. While there is no need to harden the film in 
question for the sake of toughening it, inclusion of the hardening 
constituent in the bath has been found desirable to make drying 


The spray system for applying wash water was essentially similar 
to that used in the developing and fixing tanks and used water at a 
rate of about 1.5 gpm. At a wash- water temperature of 120 F with 
the film in question, hypo and silver residues reached the level some- 
times referred to as "commercial" in 5 sec and as "archival" in 10 
sec or less. 


When rapid drying is to follow, loose water must be removed more 
uniformly and completely than in normal processing. Spots and 
streaks produced in drying when squeegeeing is inadequate cannot 
be prevented by the liberal use of an efficient wetting agent in the 
wash water, presumably because the rate of redistribution is too 
slow to keep up with the needs in the two or three critical seconds of 
the drying process. Of necessity, reliance was placed, therefore, on 
the liberal use of compressed air. About 30 cfm (atmospheric) were 
used at 15 psi. 


Measurements of the water content of Eastman Fine Grain Release 
Positive Film processed in the manner described here have shown 
that the water absorption is ordinarily less than in conventional proc- 
essing and that almost all of it is by the emulsion layer. 

The drying equipment used in this work is designed to hasten the 
evaporation of water mainly from the emulsion surface while the 
film support lies in contact with the smooth chromium-plated drum 
shielded from air circulation. No provision is made for application 
of heat except from the emulsion side. In the first experimental 
work, drying air at room temperature was supplied in forceful cross 
streams from orifices of 0.040 in. in diameter near the edge of the film 


at intervals of one-half inch on either side. Considerable depend- 
ence was placed upon the supplementary effect of radiant heat pro- 
vided by the near-by fluted Nichrome ribbon which was operated 
near the glow point, i.e., about 1000 F. With this open radiator, 
the machine was suited only for use with safety film. 

Previous studies had shown that for proper use of radiant heat the 
flow of air over the film surface should be sufficient to prevent any 
large elevation of the temperature of the film if severe physical 
effects were to be avoided. In the present work, therefore, an ample 
air stream was used to hasten the drying so that the condition of the 
film was good except for a small and unimportant increase in brittle- 
ness. Nevertheless, the proximity to the threshold of physical 
change was indicated by a tendency to increased glossiness of the 
emulsion surface. Any substantial increase in air flow was imprac- 
tical because of the loss in total efficiency of the system caused by the 
cooling of the bare radiant ribbon by the air deflected back against 
it from the film and the air distribution piping. 

More intensive air impingement with a new distributor was then 
tried with success. In place of the original cross-flow system, a lad- 
derlike structure of tubing was installed at a distance of about one- 
half inch from the film. The four 0.040-in. orifices in each rung were 
staggered relative to the width of the film. With 40 rungs along the 
7.5-ft film length, for example, 40 cu ft of air at about 160 to 170 F 
and at 10 Ib manifold pressure were required for the Fine Grain Re- 
lease Positive Film. In general, lower temperatures are preferred, 
with a greater number of rungs delivering a proportionately increased 
volume of air. An improved design is now being built for the sys- 
tematic study of temperature, pressure, orifice diameter and spacing. 
Preliminary data indicate that groups of orifices of the type men- 
tioned here, located at intervals of 1 in. or less and delivering a total 
of 100 cu ft of air at 90 F, will be sufficient. 

Cabinet Dryer 

When, for any reason, the drying path is increased beyond 10 ft, 
the drum-type dryer is too cumbersome and will be replaced by a 
cabinet dryer in which less forcible air streams will be applied over a 
proportionately greater length of less firmly supported film. A cabi- 
net dryer of this type is illustrated in Fig. 6. 

In the dryer, the film travels, emulsion side outward, in the usual 
helical path around upper and lower rollers of the rack. Drying air 
is delivered perpendicularly to the emulsion surface from the supply 
plenum through a very large number of staggered small orifices or 




through numerous narrow slits, each covering the width of the film 
strand. Pressure is maintained in the plenum by a blower whicri 
takes air from the vicinity of the film strands and fresh air from a 
dampered intake pipe. Air leaving the blower flows through the 
thermostatically controlled heater and then enters the several sec- 
tions of the plenum. 

The slits or groups of orifices should be spaced in such a way that 
the film in each strand passes them at a rate of 15 or more a second. 
Air velocities should be upward of 100 fps. The dimensions of the 
openings will depend upon the 
supply pressure. The total 
volume of air necessary for dry- 
ing Eastman Fine Grain Posi- 
tive Film at 120 F will be of 
the order of 75 to 100 cfm, 
measured at atmospheric pres- 

The essential feature of the 
cabinet as well as of the drum 
design is the frequent sweeping 
of the emulsion surface by 
forcible streams of unsaturated 
warm air. During the 5 sec 
the film is in the drum dryer, 
the force of the air blast is so 
great that the film must be 
supported rigidly at all times. 


Film (Door Removed) 

*. indicotes 
surfaces having 

Fig. 6. Cabinet-type rapid film- 
dryer employing numerous jets of 
high-velocity, heated air. 

When 10 to 20 sec of time are 
available, sufficient support 
may be provided by a few 
backing rollers or by balancing air jets applied to the film support. 

By the use of the methods and equipment discussed here, the re- 
quirements of theater television and the like for simplified, auto- 
matic processing of Fine Grain Release Positive Film in a restricted 
space can be met. Depending upon the requirements for quality and 
permanence in a given case, the time of processing, including drying, 
may be reduced below the 25 to 40 sec employed in the practice de- 
scribed here. 


At present, rapid processing equipment might be adopted in com- 
mercial laboratory work because it requires less space and entails 
smaller capital outlay than conventional equipment. Justification 


might be on this basis where additional equipment is needed for a 
specific task during a limited period of time. In special situations the 
simplified temperature control requiring no refrigeration, which is 
enjoyed when the processing temperature is well above ambient, 
will be of importance. Occasionally, equipment and a method of this 
type will be valuable because the delay in processing before a short 
length of film is available for subsequent use is reduced. 

In many of these cases, reduction of the processing time below 2 or 
3 min would not be necessary and greater running speed even with 
proportionately larger size might be desirable. A longer film path 
could be adopted for a motion picture laboratory machine in which 
highest quality of results is of prime importance and to permit the 
use of more dilute baths, possibly in a cascade flow through two tanks 
for economy. With more time for washing, savings could be effected 
by heating water only to 70 F. 

Fig. 7. Straight-line tube equipment for ultrarapid processing. 


An extreme case is that in which the exposed film, in a continuous 
length, must be made to produce a visible image at the earliest 
moment after leaving the exposing station. For this purpose, in 
which the film can be used directly without washing or drying, a 
straight-line machine of the type illustrated in Fig. 7 was devised by 
one of the authors about 1937. It consisted of a jacketed tube 2 ft 
long separated into three compartments by means of sponge-rubber 
plugs held in place by friction with the tube wall. The device was as- 
sembled with the leader film passing in a straight line from one com- 
partment to the next through knife cuts in the sponge-rubber plugs. 
The end compartments about 8 to 9 in. long were filled with de- 
veloper and fixing bath, respectively, while the smaller space in the 
middle was empty. The exposed film was attached to the leader and 
drawn through the baths at the rate of about 100 fpm so as to provide 
about one-half second in each bath. The highly hardened, low-speed 


Kodalith Type film was used with the strongly alkaline Kodak De- 
veloper D-9 and an ammonium thiocyanate fixing bath at about 150 
F. The latter was made up in the proportion of 15 g of the salt to 
5 ml of water and solidified at room temperature. Equipment of this 
type is, of course, very limited in application but it gives some indi- 
cation of the possibilities when the rapid processing methods and 
equipment are properly chosen for a special purpose. 


Up to this point the practice of rapid processing has been treated 
mainly for the case in which the film to be used is so highly hardened 
in manufacture that it can be subjected to severe chemical treatment 
at high temperature without causing much swelling or softening of 
the gelatin emulsion layer. Among the motion picture films regu- 
larly supplied only certain of the lower-speed types are hardened to 
this degree during manufacture. In addition, a few special photo- 
graphic films have been made for applications where certain limita- 
tions in properties and restrictions in handling can be accepted. At 
the present time it is not possible to make commercially satisfactory 
high-speed negative emulsions hardened to this degree, although prog- 
ress is being made. 

In order to extend the benefits of the rapid processing procedure to 
the emulsion types which are not available fully hardened, modified 
techniques have been studied in which supplementary hardening is 
given at the start of processing or in which the severity of the treat- 
ment is moderated with some concession in length of treating time. 

Method with Prehardener 

The use of a prehardening bath, such as Kodak Prehardener SH-5, 
is satisfactory with most types of high-speed negative films and per- 
mits subsequent development at temperatures up to 125 F, but con- 
sumes from 1 to 4 min in various cases. However, this treatment 10 re- 
moves all obstacles to the use of the strenuous rapid processing treat- 
ment without causing any significant loss of emulsion speed or image 

From the prehardener the film can go directly to a vigorous de- 
veloper, such as Kodak Rapid Developer SD-26. This is followed by 
rapid fixing, washing and drying procedures of the type discussed in 
the preceding sections. Times of treatment for the higher-speed 
negative materials must be appropriate for the combination of film 
type and processing baths chosen but will usually be several times as 
long as for Fine Grain Release Positive Film. 


When it is imperative that the time of prehardening be reduced, the 
temperature in this prebathing can be elevated to 125 F if the SH-5 is 
modified by the addition of 100 g of anhydrous sodium sulfate, and 
45 ml of formalin per liter. High-speed negative films can be fully 
hardened in this bath in 30 sec to 1 min, but when so treated will 
yield only about one half the normal emulsion speed. The anti- 
foggant concentration may require adjustment for optimum emul- 
sion speed with a given combination of film type and developer. 

Intermediate Method 

A preferable scheme of handling the less hardened films including 
the high-speed negatives is to employ all the features of the rapid 
processing technique such as the use of rapidly acting baths, the 
spray bathing and washing, and impingement warm-air drying, 
avoiding only the use of high temperatures. In this way, the high- 
speed negative films can be processed completely at 70 F to give good- 
quality images in 4 min, that is, in one tenth the usual time, by the 
use of Kodak Rapid Developer SD-26 (formula below), Kodak 
Rapid Liquid Fixer (with Hardener), spray washing, and impingement 
warm-air drying, each for 1 min. A 2-sec spray rinse with water is 
sufficient between developing and fixing. 

Kodak Rapid Developer SD-26 

Water, about 90 F (32 C) 750.0 ml 

Kodak Elon Developing Agent 20 . g 

Kodak Sodium Sulfite, desiccated 60 . g 

Kodak Hydroquinone 20 . g 

Kodak Sodium Hydroxide (Caustic Soda) 20 . g 

Kodak Potassium Bromide 10 . g 

Cold water to make 1.01 

With the drying air moving at high velocity over all parts of the 
film, the temperature of the wet emulsion approaches the wet-bulb 
temperature which is below 80 F for a dry bulb of 120 F, even when 
the air is taken into the system at 70 F and 70% relative humidity. 
A further margin of safety against softening of the emulsion can be 
obtained where required by an increase of 1 min in the time of treat- 
ment in the hardening fixing bath, which should be obtained prefer- 
ably by the addition of a second fixing-bath cabinet, into which the 
replenisher bath is fed in a two-stage counterflow system. 

Composition of Processing Baths 

The formulas given here have been found useful in practical ap- 
plications but will require modification to suit the needs of individual 


film types, the limitations and peculiarities of equipment, and to meet 
the chemical and economic requirements of replenishment and silver 
recovery. Additional information on the chemistry of rapidly act- 
ing baths will be found in a series of papers by J. I. Grabtree and his 
associates 11 " 14 on rapid processing and on low- and high-temperature 

Because of the short times and intense agitation used in practice, 
preliminary tests with hand manipulation are of limited assistance in 
selection of chemical bath formulas, and should be followed by tests 
on a typical element of the machine design under consideration be- 
fore final decisions are made. For example: the characteristic 
curve may show a drooping shoulder with quiet immersion develop- 
ment; high fog may be caused by excessively slow transfer from 
developer to rinse, insufficiently rapid renewal of rinse water, or lack 
of agitation in rinsing; yellow stain which may be difficult to elimi- 
nate in hand tests without use of an acid stop bath is easily over- 
come in the machine by forceful spray rinsing with water. 

An unusual characteristic of the current variable-density sound- 
recording film was observed when high-activity hydroquinone or 
Elon-hydroquinone developers were adopted for rapid processing. 
When development in these baths was carried to the point where the 
normal low-contrast curve was obtained with low-intensity exposures, 
a contrasty continuously upcurving characteristic was found with 
high-intensity short time exposures such as are used in sound-record- 
ing or in kinescope photography. The effect was observed at 70 F as 
well as at higher temperature. Normal curve shape was obtained 
by the addition of 10 g of sodium thiosulfate per liter to the rapid 
developer. The effect of exposure intensity level on curve shape with 
normal developers is very small. 

It has been supposed that excessive consumption of developer 
might occur with spray application of warm developer. In practice, 
this is not a serious problem since the air in a small developer cabi- 
net is insufficient to oxidize any large amount of sulfite, and renewal of 
air can be kept small by the use of the tight seals which are required 
for other reasons. Nevertheless, troublesome aerial fog was encount- 
ered in one case, even when extra precautions were taken to reduce 
the amount of air leakage. This difficulty was eliminated when the 
developer alkalinity was lowered a few tenths of a pH unit below the 
critical point for aerial fog propensity near 12.0 found by H. D. 
Russell and M. D. Little, of these Laboratories (private communica- 

In connection with the increasing use of spray application of de- 


velopers, the relation between features of design of a chemical re- 
circulation system affecting aeration, and the economics of the de- 
veloping agent and sulfite consumption have been studied by G. I. P. 
Levenson. 15 He concludes that serious losses can occur when air is 
introduced into a developer by spraying or other means, especially if 
the volume of developer in a recirculation system is large. His find- 
ings indicate the desirability of extremely small circulatory systems 
relative to the rate of film handling, as exemplified by the spray de- 
veloping unit described in the present paper. 

Effect of Temperature Elevation 

As stated by Crab tree, 13 the rate of development generally in- 
creases by a factor of about 2 for each 15 F rise in temperature. 
The acceleration of fixing by elevation of temperature is much less. 
Washing of film can be speeded up greatly by rapid renewal of water 
at the film surface but the influence of temperature elevation, while 
favorable, requires further study. Unnecessarily high temperature 
of the wash water should be avoided, both to prevent swelling and 
softening of the gelatin emulsion and to economize power in water 
heating. The effect of temperature variation in rapid processing 
has proved to be about the same per degree as in conventional 70 F 

Quality of Rapidly Developed Images 

Up to the present time, no systematic study has been made of the 
effect of rapid processing on the structure of the developed silver 
image to discover the effect on resolution, graininess, etc. However, 
observation on images developed, fixed, washed and dried in times of 
5 to 10 sec, respectively, by projection and in photomicrographs has 
shown little that is unusual. In certain cases, evidence has been ob- 
tained of incompleteness of treatment near the bottom of the emul- 
sion layer with a development time of 5 sec even though the emulsion 
speed and quality were about normal. This deficiency has been found 
to increase when the treating time was reduced to 1 second, for ex- 
ample, especially if the compensatory adjustments in developer ac- 
tivity and temperature were not sufficient to assure normal complete- 
ness of development. The use of elevated temperatures appears to 
offer no promise of a significant increase in emulsion speed nor of 
improvement in graininess. Anyone contemplating the use of high- 
temperature processing with the softer types of emulsion should make 
sure that the supplementary hardening in processing is always ample 
so that graininess will not be produced as a result of incipient reticula- 


Mechanical Condition of Processed Film 

A marked embrittlement of a type which is not removed by equili- 
bration with an atmosphere of 70% relative humidity is observed oc- 
casionally when radiant heat is used improperly in drying. The 
condition appears to be caused when the film reaches an excessively 
high temperature in the course of drying and can be largely corrected 
by rewetting the film and drying it under more favorable conditions. 
It is not a direct consequence of rapid drying but rather of overheating 
during drying. 

Almost all of the heat which goes into the drying process is ac- 
counted for as the latent heat of evaporation of the water so that, 
unless evaporation is retarded by the accumulation of moisture vapor 
above the film surface as a result of insufficiently high velocity of the 
drying air, a high rate of heat input produces only a moderate rise in 
temperature of the film. Therefore, a first concern in designing a 
drying system is to have proper velocity and distribution of the air. 
Only when this is assured is it safe to introduce the large amount of 
heat which will be required for rapid evaporation of water. 

Effect of Rapid Processing on the Film Support 

Up to the present time, no detrimental effects of rapid processing 
on the film support have been observed when temperatures and 
physical handling were reasonable. As the temperature actually 
attained by the film is raised above 125 F, additional care is required 
to limit the tension applied because the film becomes more susceptible 
to plastic deformation. 


Some of the advantages obtained by the use of the rapid processing 
technique are as follows: 

1. Elimination of delay in obtaining completed film. 

2. Simplification of equipment design, which affects construction, 
maintenance and use. 

3. Reduction of volume of baths, especially if spray application is 
used, thereby reducing the amount of space required and simplifying 
chemical control. 

Some of the disadvantages of the methods considered are : 

1. Increased chemical consumption in certain cases where con- 
centrated baths are replenished rapidly. 

2. Increased power consumption when certain extreme require- 
ments as to operating characteristics or size are imposed. 



It is apparent that when the operations in processing are shortened 
by a factor of 10 to 50 times or more, attractive possibilities are of- 
fered. However, there is no universally best design for equipment 
to be used with rapid processing methods. Instead, the design must 
be chosen in any particular case according to whether the emphasis 
is to be on compactness, curtailment of the time of processing, sim- 
plification of operation, etc. General comments on design of equip- 
ment are given in Appendix I and details in regard to components in 
Appendix II. 


The authors are glad to acknowledge the valuable contributions made to this 
work by Norman A. Exley, of these Laboratories, particularly in the work on the 
"intermediate method" by which the softer high-speed negative films can be 
processed rapidly without the use of a prehardening bath. They also desire to 
express their thanks for the generous assistance in problems of equipment design 
given them by the engineering personnel of the Eastman Kodak Co. 

Rochester 4, N.Y. 
September 28, 1949. 


1. Film Path. In applications where unskilled operators will use 
the equipment, threading should be extremely simple, or else fully 
automatic. However, unless an ultrarapid process is employed, 
the longer film-path length of fast-running machines must be sinu- 
soidal or helical for the sake of compactness. In order to cut down 
the horizontal length of the machine and the number of film-trans- 
port rollers, the film loops could be lengthened, but when the span 
between supporting rollers exceeds about 5 ft it becomes increasingly 
difficult to keep the film in position under the action of forceful jets. 
The use of drums is limited practically by their bulk, even if the film 
makes more than one turn around them. 

2. General Arrangement. With equipment such as the 90-fpm unit, 
a turn-around path could be used to place the start and finish, side by 
side. The machine could then be located in an alcove on a roll-away 

3. Number of Stages. As stated previously, the provision of more 
than a single stage in the bathing treatments where space restric- 
tions are not extreme gives desirable latitude in formulating baths 
and makes possible more economical operation. 


4. By the use of stepped-roller shoulders and the interposition at 
suitable points of fully supporting soft-rubber rollers, . 16- and 35- 
mm film can be run alternately without pause. 

5. To avoid excessive carry-over or dilution of the concentrated 
solutions which would otherwise occur in the necessarily very short 
cross-over paths, soft-rubber wringer rollers or other squeegee devices 
must be used. 

6. Film Drive. In the equipment described, the film receives 
positive drive at a single roller near the wind-up. This is made 
possible by the shortness of the film paths and by the reduction in 
film-dimensional change in the shortened processing times. As the 
length of the film path and the number of rollers is increased to pro- 
vide for additional stages of treatment, it is to be expected that pro- 
vision for relief of accumulated film tension will be required. 

7. Spray Application. Forceful application of an over-all fine spray 
appears most practical for vertical film strands. A flooding-type 
low-pressure nozzle has the advantage of operating at low pressures 
and does not clog easily. However, it is usually applicable only to 
horizontal film paths. Widely spaced solid stream jets are usually 
not suitable. 

8. Splicing and Roll Handling. Suitable rapid automatic splicing 
is needed with fast machines to eliminate the need for bulky film- 
reservoir elevators. Developments in heat splicers show promise. 
In low-speed machines, threading can be made so simple that leaders 
can be dispensed with. The machine can be stopped momentarily for 
rethreading after the tail end of the preceding roll has been allowed to 
run through. 

9. Power Consumption. It should be possible to increase the ef- 
ficiency of squeegeeing immensely by improvement of the pneumatic 
method or by introduction of another. Likewise, with the use of air 
recirculation and with low-pressure fans in place of compressors, 
power consumption during drying can be minimized. Machines 
designed for greater economy of power and water consumption with 
provision for removal of hypo from water 16 and for daylight opera- 
tion should prove useful for military and other uses in which port- 
ability is required. 


1. Thermometers. When the solutions are rapidly circulated, the 
temperature can be measured accurately by means of an industrial- 
type thermometer properly located in the spray-nozzle feed lines. 
Examples are the Weston Dial Thermometer Model 221-D, Range 


0-200 F, and the Rochester Manufacturing Co. Model No. 1758, 
Range 0-200 F. 

2. Thermostats for Liquid Solutions. In the circulatory systems 
described, the simplest types of on-off thermostats have been used 
successfully when located immediately downstream of low-lag type 
heaters. Industrial-type thermostats with thin stainless-steel pro- 
tective wells, such as the Fenwal Thermoswitch A-7100 (well extra), 
can be used. In differently arranged systems or in case equipment 
is to be operated with both refrigeration and heating elements or for 
long periods without checking, it is likely that more elaborate con- 
trols will be required. Tempered water can be obtained most con- 
veniently by the use of thermostatic mixers, such as the "HE" series 
furnished by Powers Regulator Co. 

3. Solution Healer. Highly responsive electric resistance heaters 
sheathed in stainless steel and fitted for insertion in threaded pipe 
are well suited to use in the small recirculating systems. For ex- 
ample, in the 90-fpm machine, a 4,000-w multiple-connection Screw- 
In Immersion Heater, Model No. 3-887, supplied by the American 
Instrument Co. with U-shaped heating coils, provides for rapid 
heating in the warm-up period and then operation with lower wattage 
when the 120 F level has been reached. At an ambient temperature 
of 70 F and with 70 F replenisher flowing in at the rate of 300 ml a 
minute, a 1000-w heater must operate almost continuously to main- 
tain temperature at 120 F in the 90-fpm machine described. The 
switching relay can be made to rearrange the load connections auto- 
matically in proper sequence. 

4- Solution Heat Exchanger. The heat exchanger can be assembled 
from standard stainless-steel pipe and fittings chosen to suit the 
length of the heating coils and to assure full velocity of the circulated 

5. Solution Filters. All spray systems should be equipped with 
strainers of 50- to 100-mesh wire screen just upstream of the dis- 
tributing manifold to prevent clogging of nozzles. Stainless-steel 
cloth is available for this use in any degree of fineness needed. A 
coarser screen should be installed in the outlet from the sump to pro- 
tect the pump. A filter, such as the Fulflo (Commercial Filters 
Corp.) Model ABR-8, but specified in stainless steel instead of the 
usual brass, can be used with the disposable 8-in. filter cell, 1541 
KCOSS-1, which consists of cotton fiber wound on a stainless-steel 
wire-mesh core. However, the need for a filter should be established 
relative to the type of use because the rapid purging in these small 
systems minimizes sludge accumulation. 


6. Spray Nozzles. The forceful solid-cone spray required for 
washing and usually for solution application can be obtained by the 
use of commercially available nozzles, such as the stainless-steel 3^ 
GGSS-1 supplied by Spraying Systems Co. At 30 psi, this nozzle 
is rated to deliver 0.17 gpm. A diversity of nozzle types are avail- 
able from several manufacturers and can be obtained in various other 
materials, such as Monel metal, and hard rubber. In the 90-fpm 
machine, nine nozzles were sufficient to cover the film in two loops 
totaling 15 ft in length, while five additional nozzles were needed for 
an additional loop. 

7. Pumps. Several types of pump suitable for supplying the 2 gpm 
at 30 psi required in the case of the 90-fpm machine are available. 
Corrosion resistance equivalent to that of "18-8 molybdenum" 
(American Iron and Steel Institute), Type 316 or 317, is desirable, 
especially with fixing baths, while Type 304 is suitable for developer. 
Other highly resistant metal, glass, or plastic materials may be re- 
quired for use with more corrosive baths, such as bleaches, toning 
baths, etc. 

8. Drying Air Temperature Control. Drying systems in which the 
air is recirculated present the usual problems of regulation and may 
require hygrostats as well as thermostats. In contrast, once-through 
systems which take low-temperature room air and raise its dry-bulb 
temperature 30 F require only a thermometer and means for manually 
setting the power input to the heaters if the quality of the supply air 
is constant. Problems of regulation as well as operation are entailed 
if the air distribution system has a large thermal capacity or slug- 
gish heaters. Rapidly responding electrical heaters are generally 
most suitable and should be placed close to the discharge orifices. 

9. Air Heaters. A wide variety of electrical heaters are avail- 
able in forms suitable for the space in which they must be installed. 

10. Rollers. Because they were readily available, molded rollers 
of "non-fogging" Bakelite were used in the 90-fpm machine, even 
though they were eroded rather rapidly by the hot high-pH developer. 
Rollers were turned from other more resistant plastics or even from 
stainless steel for some of the other machines. 

11. Corrosion-Resistant Construction. In some cases, it will be 
impractical to solve the problem of corrosion resistance by specifying 
construction with certain materials. It may then be necessary to 
plan to renew certain parts from time to time. Considerable assist- 
ance can be obtained from the recently published Corrosion Hand- 
book 11 and from the booklet, "Materials for the Construction of Photo- 
graphic Processing Apparatus." 18 



1. E. Goldberg, "Anwendung der Fernsehmethoden in der photographischen 
Technik," Kinotechnik, vol. 14, pp. 401-402, 1932. 

2. G. Schuberg, W. Dillenburger, and H. Zschau, "Das Zwischenfilmverfahren," 
Fernseh G.m.b.H. vol. 1, pp. 201-210, Dec. 1939. 

3. A. G. D. West, "Television in the cinema," To-Day' 's Cinema, pp. 5-9, June 
27, 1935. 

4. F. E. Tuttle and C. H. Green, "Photographic race timing equipment," 

Jour. SMPE, vol. 27, pp. 529-536, Nov. 1936. 

5. A. R. Burkin, "Ultra-rapid processing of photographic materials," Phot. 
Jour., vol. 87B, pp. 108-111, Sept.-Oct. 1947. 

R. J. Hercock, "The photographic recording of cathode-ray tube screen 
traces; the choice of emulsions and developers," Phot. Jour., vol. 86B, 
pp. 138-142, 1946. 

6. F. Brown, L. L. Blackmer, and C. J. Kunz, "A system for rapid production 
of photographic records," /. Frank. Inst., vol. 242, pp. 203-212, Sept. 1946. 

7. B. K. Thorne, "Air Force 'Brass' Gets 'At Ease' Sky," New York Times, 
June 27, 1949. 

8. D. S. Bond and V. J. Duke, "Ultrafax," R.C.A. Review, vol. 10, pp. 99-115, 

Mar. 1949. 

9. J. G. Capstaff, Eastman Kodak Co. U.S. Patent No. 2,289,753, 1942. 

10. H. A. Miller, J. I. Crabtree and H. D. Russell, "A prehardening bath for 
high-temperature processing," /. Phot. Soc. Amer., vol. 10, pp. 397-404, 
Sept. 1944; vol. 10, pp. 453-458, Oct. 1944. 

11. H. Parker and J. I. Crabtree, "Rapid processing methods," Jour. SMPE, 
vol. 26, pp. 406-426, Apr. 1936. 

12. J. I. Crabtree and H. D. Russell, "Rapid processing of photographic mate- 

rials," J. Phot. Soc. Amer., vol. 10, pp. 541-551, Nov. 1944. 

13. R. W. Henn and J. I. Crabtree, "Photographic processing at low and sub- 

zero temperatures," /. Phot. Soc. Amer., vol. 12, pp. 445-451, Oct. 1946. 

14. J. I. Crabtree, "Rapid processing of films and papers," /. Phot. Soc. Amer., 
vol. 15, pp. 130-136, Feb. 1949. 

15. G. I. P. Levenson, "The chemical economics of spray processing," Brit. Kine- 
mat., vol. 14, pp. 65-81, Mar. 1949; also G. I. P. Levenson, Jour. SMPE, 
vol. 53, pp. 665-690, Dec. 1949. 

16. H. P. Gregor and N. N. Sherman, "Demineralization of photographic wash 
water by ion exchange," Jour. SMPE, vol. 53, pp. 183-192, Aug. 1949. 

17. Corrosion Handbook, H. H. Uhlig, Editor, J. Wiley & Sons, New York, 1948. 

18. J. I. Crabtree, G. E. Matthews, and L. E. Muehler, Materials for the Con- 

struction of Photographic Processing Apparatus, Eastman Kodak Co., Ro- 
chester 4, N.Y. 

A 16- Mm Rapid Film Processor 





SUMMARY: The proven practicability of spray processing, coupled with the 
availability of acetate film bases which will withstand fairly high processing 
temperatures, enables construction of compact continuous processing equip- 
ment for operation at synchronous speed of cameras and projectors. The 
theory and construction of an experimental equipment are described. Sig- 
nificant performance features are studio print quality, continuous automatic 
operation and convenient control of process variables. Auxiliary equipment 
permits reel-to-reel, camera-to-projector or camera-to-reel processing. 
Possible applications are television network service hi connection with video 
recording, motion picture theaters, laboratory processing of small film 
batches, and motion picture studio monitoring of critical takes. 

ALTHOUGH PRODUCTION of release prints in 16-mm size has in- 
creased considerably in the past few years, very few users of 16- 
mm film have been able to maintain complete processing facilities for 
preparing their own prints immediately after photography. Recent 
developments in film bases and emulsions now enable construction of 
compact high-temperature, continuous processing equipment for this 

Continuous film processors operate on fundamentally the same 
principle, regardless of their size or their speed of operation. The film 
travels at a steady rate through a series of processing chambers, where 
it is developed, rinsed, fixed, washed and dried in accordance with a 
definite time cycle. Time allotments in the wet stages of the cycle 
are determined largely by solution strengths and temperatures. In 
the drying stage, the time allotment is determined by the film water 
content and -by the effectiveness of the drying method. The sum of 
the separate time allotments is the total processing time. The inter- 
nal film path is sufficiently long to allow completion of the processing 
cycle while the film travels through the machine at a predetermined 

The physical form of a processor is, however, subject to wide varia- 
tions, depending on the operating condition for which the particular 
processor is designed. Commercial bulk film processors are designed 
for high production quotas, involving film travel rates of at least 

PRESENTED: April 25, 1950, at the SMPTE Convention in Chicago. 





150 fpm. These machines have long internal film paths and use 
large quantities of developer and fix solutions. They are therefore 
quite large, requiring one or more rooms for complete installation. 
Threading is a lengthy operation. The travel time from one end of 
the machine to the other may be as long as 30 min. 

Television film processing introduces a new operating condition, 
calling for a different type of processor. In this case, the processor 

Fig. 1. Rapid film processor, front view. 

receives film directly from the camera, and the prime requirement is 
that the film should be ready for projection in as short a time as 
possible. The machine developed for this purpose is called a Rapid 
Film Processor. It differs from a bulk film processor in several 
respects. The film travel rate is only 36 fpm, which corresponds to 
the average rate of 16-mm film, exposed at 24 frames/sec. Since the 
rate is much lower than that required in a bulk film processor, the 




internal film path can be much shorter, threading can be simpler, 
and construction can be much more compact. Solution quantities 
are correspondingly smaller, and hence there need be no major loss of 
time and chemicals when starting and stopping the processor. Special 
measures are, however, required to shorten the processing time in both 
the wet and dry stages. 

The introduction of hardened emulsion type film (Eastman Fine 

Fig. 2. Rapid film processor, rear view with covers closed. 

Grain Release Positive, Type 7302) permits rapid processing at ele- 
vated temperatures. A rule-of-thumb formula indicates that proc- 
essing time is cut in half for each 15 F (degrees Fahrenheit) tempera- 
ture increase. 

The film stock is intended for positive prints, but can be used as a 
negative material in noncritical applications. It is low in cost and 
has the added advantage of a fine grain emulsion. 


The rapid film processor which will be described is an experimental 
model. It was constructed primarily for evaluation of space require- 
ments, control features, and film receiving and discharge methods. 
The flow rates, pressures and temperatures employed in this unit are 
based on those used in a 35-mm pilot unit which was built by Eastman 
Kodak Co., under the direction of C. E. Ives and C. J. Kunz (see 
pp. 3-26 of this issue of the JOURNAL). Controls, indicators and 
measuring devices have been planned to permit further experimental 
studies. Simplification of controls will be necessary and desirable 
for commercial production. 

The experimental processor stands 5 ft high, 2 1 /z ft deep, and 3 ft 
wide, exclusive of side film storage compartments. It produces fin- 
ished film, ready for projection, in 40 sec from the start of processing. 
Print quality is comparable to that obtained in larger machines oper- 
ating on a much longer processing cycle. The film is thoroughly 
washed and fixed, and has a sufficiently low hypo content for long- 
term or archival storage. 


Front and rear views of the processor are shown in Figs. 1 and 2. 
The film is carried on reels in storage compartments in the side covers. 
Exposed film from the right storage compartment is processed as it 
passes through the console, and is taken up as finished film on a take-up 
reel in the left storage compartment. With different film routing 
through the storage compartments, film can be received from a camera 
film tunnel, or can be delivered directly to a projector. Switches are 
provided to control a camera and projector, starting or stopping them 
simultaneously with the processor film drive. 

The film travels on spools through the three processing tanks 
shown in Fig. 3. Spray processing is used in all three tanks to avoid 
directional effects encountered in dip processing. These effects are 
caused by diffusion of development products from exposed film areas 
onto adjacent film areas that have had different exposure. With spray 
processing, a uniform solution concentration is delivered to all film 
areas, and development products are continuously removed. The 
solution penetrates deeply into the film emulsion. The exposed film 
passes first through the developing tank, then through the rinse and 
wash tank to the fix tank. The film then returns to the rinse and 
wash tank, from where it travels upward through the air squeegee 
into the drying chamber. 

The air squeegee, in effect, scrubs the film with a stream of pre- 
heated air, bodily removing surface water. The film leaving the air 




squeegee requires further drying, but has no surface droplets to spot 
the base or the emulsion. The film then passes through a drying 
chamber where infrared lamps and circulating air complete the drying 
process. After drying, the film is ready for waxing and projection. 

The processing solutions, and the rinse and wash water, are used at a 
normal temperature of 120 F. The rinse and wash water is discarded 
after a single pass through the processor. The processing solutions 
are conserved through recirculation, and are replenished at a constant 
rate to maintain working strength. The spent solution residues mix 
with the rinse and wash water in the sump of the processor, where 
they neutralize each other almost completely and become sufficiently 
diluted for disposal in a sewage system. 


v -RINSE a 



Fig. 3. Interior arrangement. 

High solution temperatures and effective drying techniques reduce 
the total film processing time to 40 sec between the input and output 
ends of the processor. This time is divided as follows: develop, 
5 sec; rinse, 2 sec; fix, 10 sec; wash, 5 sec; dry, 15 sec; and inter- 
process film transport, 3 sec. 

The film drive sprocket (Fig. 3) is the only point in the processor 
where film is positively driven. The synchronous motor which 
drives the sprocket.also drives the spindles which carry the upper film 
spools in the tanks and drying chamber. The spindles rotate at a 
higher rate than the film travel rate, and the resultant drag between 
the spindles and the spools assists the film in its travel through the 


processor. This aided drive limits the film tension to less than 8 oz 
at any point. 

In the tanks and drying chamber, the film travels in helical closed 
loops, with emulsion side out. Multiple loops stack compactly, so 
that a single group of wide angle sprays covers all the loops within a 
tank. The tanks are large enough for convenient film threading. 

Intertank traps prevent spray transfer between tanks. The traps 
are plastic boxes with bottom drain holes, containing upper and lower 
rollers between which the film passes. The upper roller, being gravity 
loaded, rests on the film and confines the spray. An entrance trap 
seals the film tunnel against developer spray and wets the film uni- 
formly at the start of development. 

Controls for the developer system are grouped on the right side of 
the front panel, and are matched symmetrically by corresponding fix 
controls on the left side. The developer bottle, which holds 2J^ 
gal of developer, is inverted into a stainless-steel reservoir at the top 
of the processor. An air trap bottle closure of the type used in 
chicken feeders maintains constant fluid level in the reservoir and seals 
the bottle against liquid spillage during insertion and removal. From 
the reservoir, the developer flows by gravity through a needle 
valve, which controls the replenisher flow rate, and through a flow- 
meter to the pump input line. The pump delivers filtered and heated 
developer to the spray nozzles in the developing tank. The spray, 
after impinging on the film loops, falls into the sump, where an over- 
flow pipe maintains a constant level and continuously drains a portion 
of the spent solution. The remainder of the spent solution returns 
to the pump input, where it is replenished with fresh developer from 
the developer bottle and recirculated. 

Two thermostatically controlled heaters maintain solution working 
temperature within a tolerance of % a degree. The first heater, a 
coarse heater with a high rating, functions on starting and cuts out at 
5 F below the operating temperature. The second heater, a fine 
heater with a lower rating, cuts out when the solution reaches operat- 
ing temperature. 

The controls for the fix system are the same as those for the devel- 
oper system. Normal replenisher flow rate for each system is 60 
to 70 ml/min, or approximately 1 gal/hr. Two gallons are required 
for initial priming of each system. Approximately 0.9 gal/min are 
sprayed through each set of nozzles during operation. 

The controls for the rinse and wash system are grouped in the center 
of the front panel. Operation of the system is entirely automatic. 
A thermostatic mixer, mounted on the front panel, maintains accurate 


water temperature as long as the incoming hot water supply is hotter 
than the operating temperature. A panel thermometer indicates the 
hot water supply temperature and lights a warning light if necessary 
to indicate Subnormal Water Supply. The combined flow of hot and 
cold water supplies is 1J/2 gal/min. 

The air squeegee housing contains two orifice blocks which direct 
air streams onto opposite sides of the film. The film enters the hous- 
ing through a lower pair of rollers, spaced 0.008 in. apart, and leaves 
the housing through an upper pair of rollers, spaced 0.006 in. apart. 
Since the upper rollers provide no film clearance, the air stream is 
confined to the 0.001-in. clearance space under each lower roller. As 
a result, the air stream is in close contact with both sides of the film. 
The air stream bodily removes surface moisture from the film and 
carries it downward toward the sump. Each pair of rollers is spring 
loaded so that it yields to permit passage of a film splice, but does 
not deflect under normal air flow pressure. The housing may be 
opened for film threading. 

Two 500-w infrared lamps heat the film in the drying chamber, 
driving moisture out of the film emulsion. An exhaust blower at the 
top of the drying chamber carries away moisture-laden air. Clean, 
dry air enters the console through filter panels in the side covers. 


Before the film leaves the drying chamber, it passes through a dip 
bath containing a solution of carnauba wax in carbon tetrachloride. 
It is generally recognized that films which have been waxed in this 
manner are more durable than unwaxed films. The film dries com- 
pletely before it leaves the chamber, and the fumes are carried away 
by the exhaust blower. The film is then ready for projection. 

As the finished film discharges from the processor it is either 
delivered to a projector or taken up on a storage reel within the side 
cover. One experimental type of side cover carries fittings for both 
methods of delivery. In addition to a take-up-reel spindle, it has a 
film storage elevator containing an isolating film loop. The maximum 
footage which can be stored in the loop is equivalent to 10 sec of run- 
ning time. The main purpose of the loop is to permit simultaneous 
starting or stopping of processor and projector, allowing for a dif- 
ference between the separate machine rates during the transition 

A large-capacity film storage elevator is available which will per- 
mit as much as a 3-min delay between processor and projector. By 
use of this elevator, film can be monitored and edited prior to pro- 




jection. The unit can be equipped with a commercial viewer, such 
as a Craig viewer, which has been found very useful for continuous 


A series of test runs has been made at different operating tempera- 
tures. The results of three of these runs are shown in Fig. 4, in the 
form of H & D curves. All runs were made with Eastman Fine 
Grain Release Positive Film Type 7302, processed in D-8 developer. 
At a temperature of 120 F, the temperature at which the required 
increase in processing rate was obtained, a density of 3 is attained 
within the linear portion of the curve. 

5 10 15 20 

Fig. 4. Density vs. exposure at different temperatures. 

It may be noted that the upper temperature limit of the series of 
runs exceeds the allowable processing temperature of the film. The 
purpose of these runs, however, was to determine the range of gamma 
variation in processing at different temperatures. The curve re- 
produced in Fig. 5 shows a very linear variation with temperature, 
demonstrating the practicability of gamma control through tempera- 
ture setting. 

The residual hypo content of film processed at normal temperature 
(120 F) was tested by the mercuric chloride method described in 
American Standard Specifications Z38. 3. 2-1945 "Films for Permanent 
Records." The film samples for test were taken at random from 
several film batches, and were split into two groups. The samples 
of one group were washed thoroughly in carbon tetrachloride to re- 




move the lubricating coating or carnauba wax; the samples in the 
other group were untreated. All samples in both groups showed a 
sufficiently low hypo content to be well within the acceptable limit. 
On the basis of hypo content, the film qualifies satisfactorily for use in 
permanent records. 


The fact that finished film can be reviewed within a minute after the 
event has been photographed is possibly the most valuable single char- 
acteristic of the processor. 

In motion picture studio practice, special sets must often be re- 
tained intact until the film has been processed and reviewed. The 



90 100 110 


Fig. 5. Gamma variation with temperature. 

necessity for checking unusual lighting effects may keep actors and 
stage hands on location for a much longer time than required for 
photography alone. Delays of this nature can be minimized by using 
an auxiliary 16-mm camera in conjunction with a rapid film processor. 

Motion picture theaters may now photograph and process their 
own 16-mm film. This opens a potentially tremendous new field of 
application, the possibilities of which have not been fully explored. 
A 16-mm arc lamp projector is available for theater exhibition of 
films which have been prepared in this manner. 

The processor is expected to become a useful tool for industrial 
laboratories. Machinery studies are often recorded photographically 
on motion picture film, but the results are not available until the film 
has been processed. The delay entailed in commercial processing 


extends the time of a test program and increases the cost. The rapid 
film processor, on the other hand, can be used to make permanently 
recorded results almost immediately available. No major loss of 
time and chemicals is involved in starting and stopping the machine, 
and since threading is a relatively simple operation, small discontinu- 
ous film batches are readily handled. The processor is therefore well 
suited for laboratory operation. It is believed that use of the proc- 
essor will enable sizeable economies to be effected during the course 
of a test program. 

Films which are prepared for television broadcast by the medium of 
video recording require very close control in all phases of preparation. 
The contrast and density of the finished print are of utmost impor- 
tance. When the station equipment includes a rapid film processor 
the studio crew can control all phases of photography, including the 
recording, processing, and reproduction of the film. Factors affecting 
either contrast or density can be partially or completely corrected 
within the studio, before the film enters the projector. 


The rapid film processor is sufficiently compact for general use in pro- 
jection booths. It provides continuous automatic operation and en- 
ables convenient control of process variables. As indicated by the 
performance results, it produces film of adequate contrast, density 
and permanence to meet critical studio requirements. 


A development of this nature represents the combined work of many individuals. 
The authors wish to express to the following their appreciation for technical engi- 
neering data: H. E. White and E. Warnecke of Eastman Kodak Co. and J. G. 
Stott, formerly of Eastman Kodak Co. and now of Du-Art Film Laboratories. 

A Method of Measuring Electrification 
Of Motion Picture Film 
Applied to Gleaning Operations 



SUMMARY: A dielectric, such as photographic film, becomes electrified 
when rubbed or passed over rollers. The electrostatic charges which are 
generated attract dust and dirt particles to the film. Since dirt is objection- 
able to both the manufacturer and the user of film, means are sought to re- 
duce electrification. This paper describes a method that has been devised to 
evaluate roller-film combinations electrostatically. Film is brought to a 
given potential, either positive or negative, and the change hi potential meas- 
ured as it passes over a test roller. Typical data for a variety of rollers are 
presented. The work was extended to test the effect of rubbing film with 
cleaning pads of velvet and mouton fur. Measurements were also made 
with solvents applied to a velvet cleaning pad. 

A DIELECTRIC MATERIAL, such as photographic film base, becomes 
electrified when rubbed against almost any object, or when it 
passes over a roller, either of dielectric or of metallic composition. 
One of the effects of this electrification, and one which causes a great 
deal of trouble, is the electrostatic attraction of dust particles to the 
film. An attempt to clean the film by brushing or rubbing usually 
results in higher charges which further increase the difficulty of remov- 
ing the dust. This problem is serious to the manufacturer of the sup- 
port who seeks to produce a dust-free film, and to the laboratory 
technician who handles processed films, particularly when there is dust 
on negative film in the printing process. If emulsion-coated films 
become electrified beyond a critical value, discharges occur and fog 
and other markings are produced. 

In the handling of photographic products, one aim is to select ma- 
terials and to design equipment which will produce a minimum 
amount of electrification. Naturally, some means of evaluating 
these materials in contact with different types of photographic films is 
necessary. This is true also in the selection of the cleaning materials 
and the methods of their use. The purpose of this paper is to describe 
a method for determining the electrification properties of roller-film 
combinations. In addition, it will be shown how the method can be 

PRESENTED: April 28, 1950, at the SMPTE Convention in Chicago. 
Communication No. 1345 from the Kodak Research Laboratories. 





adapted to measure the charging and discharging characteristics of 
cleaning pads and the influence of cleaning solutions applied to these 


The schematic arrangement of equipment used in comparing vari- 
ous roller-film combinations is shown in Fig. 1. A loop of film, about 
30 ft in length, is driven over the various rollers (7*1, r z , n, r 4 , r 5 , r 6 , r 7 ) 
and the test roller, in the direction shown by the arrows. Rollers TI 
and r 2 are insulated to reduce conduction of charge in the film from 
the test roller to the ground. The film is driven at a constant velocity 
by a synchronous motor connected to roller r 5 . The film is kept un- 
der constant tension by attaching a weight, W, to a floating roller, r 7 . 

Fig. 1. Schematic diagram 
of testing apparatus. 

Fig. 2. Schematic diagram 
of field meter. 

This insures a uniform pressure of the film against the test roller. 
The film is first brought into equilibrium with a given humidity and 
temperature, and then tested in this same atmosphere. The angle of 
wrap of the film at the test roller is normally kept at 60 deg. 

After leaving roller r% the film passes between an insulated needle, 
N, and a grounded plate. By raising the needle to high potentials, 
either positive or negative, charges of either polarity may be sprayed 
onto the film. This is similar to the scheme that is used in charging 
the belt in the Van der Graaf type of electrostatic generator. A 20- 
kv, d-c power supply, in which either the positive or the negative 
terminal may be grounded, is used to supply the needle potential. 
The grounded plate concentrates the field and increases the efficiency 
of this charging process. A shield is placed around the needle to pre- 
vent stray fields from affecting near-by electronic equipment. 




An electric field-meter, FMi (Fig. 1), is placed just ahead of the 
test roller and a similar unit, FM 2 , just following it. These instru- 
ments are of the type described by Gunn 1 and Waddel. 2 A drawing 
of the detecting unit of this instrument is shown in Fig. 2. A grounded 
two-bladed sector, A 2 , rotates in front of an insulated stationary sector, 
Aij AI is alternately exposed to, and shielded from, the electrified plane 
at the right of the figure, which, in this case, is shown to be charged 
positively. The insulated sector is connected to ground through a 
high resistance, R. When exposed to the field, a charge is induced on 
AI; when AI is shielded from the field, the induced charge flows back 
to ground through R. Rotation of the sector, A 2 , repeats this process 

-10 -8 -6 -4 -2 +2 +4 +6 +8 +IOKV 
Leaving Potential V L 

Fig. 3. Curves showing electrification 
of different film-roller combinations. 

whereby an alternating potential is developed across R. This poten- 
tial is amplified by the amplifier, A, and read on the output meter, M. 
In practice, a 10-bladed sector is rotated at a speed of 1,800 rpm. 
This produces a 300-cycle signal. With an amplifier which is peaked 
for this frequency, 60-cycle disturbances are sufficiently excluded so 
that an electronic rectifier can be substituted for the mechanical rec- 
tifier system which has been more commonly used. 3 With this in- 
strument, the electric field, due to the charges on the film, can be read. 
The field meters are calibrated in terms of volts on a uniformly charged 
plate placed in the same position as the film to be measured. By us- 
ing two of these instruments, the potentials which a given area of film 
assumes before and after passage over the test roller can be read di- 

40 H. W. CLEVELAND July 

rectly. The relation between these two sets of values can be used to 
specify the electrostatic characteristic of a given roller-film combina- 

The data are plotted on a 4-quadrant type of graph, with the leaving 
potential, V L , as a function of the initial potential, F/. The film is 
brought to any initial potential, F/, with the necessary needle poten- 
tial. The initial potentials, F/, are plotted as ordinates, with the zero 
level in the center of the chart-positive potentials above the zero line, 
and the negative potentials below. The leaving potentials, V L , are 
plotted as abscissae in a similar manner, with positive potentials to 
the right, and negative to the left of the zero line. If the data fall 
on the 45-deg line drawn diagonally through two of the quadrants 
(Fig. 3), the original voltage level of the film has not been altered in 
passing over the test roller. In the upper right-hand quadrant, if 
the curve falls to the left of the 45-deg line, the film is being dis- 
charged; if to the right, it is being charged. Similarly, in the lower 
left-hand quadrant, film will be discharged if the curve is to the right 
of the 45-deg line, and charged if at the left of this line. 


Examples of running unprocessed Eastman Fine-Grain Release 
Positive Film over four types of rollers are shown in Fig. 3. The 
emulsion side of the film contacted the rollers. A chromium-plated 
metal roller or a conducting rubber roller will reduce negative poten- 
tials on the film, but will add potential to the film if it is at low positive 
potential. The resulting value for subsequent passages may be 
found by applying V L for one passage to F/ of a subsequent passage. 
By this procedure, it may be seen that the film must ultimately come 
to the potential which corresponds to the intersection point of the 
curve and the 45-deg line, e.g., for the metal roller this value is +7.5 
kv and for the conducting rubber roller +4.5 kv. 

An example will illustrate. Referring to the conducting rubber 
roller curve, if we start with an initial potential, F/, equal to about 4 
kv, the charges will be completely removed, and the leaving potential, 
VL = 0. Now, taking zero on the F/ axis, the VL value on the con- 
ducting rubber curve is +2.2 kv. Repeating this, if we apply F/ = 
+2.2 kv to the conducting rubber curve, we get V L = +3.0 kv. Con- 
tinuing this process, subsequent values are, in turn, +3.5 kv, +3.8 
kv, etc., and finally the curve intersects the 45-deg line at about +4.5 
kv. No further change in potential will take place. The film now 
leaves the test roller at the same potential it had when it reached the 


test roller. This process corresponds to the passage of the film over a 
number of rollers of this same material. 

Examination of the Lucite roller curve shows that for the film used 
in this case, a high positive potential, Vj, is always reduced by passage 
over the roller until the incoming potential, V I} becomes about +2.5 
kv. In the next passage over this roller or a duplicate roller, negative 
charges are added and the film leaves with a potential of 1.5 kv. 
Continuing the step-by-step process, the film reaches a stable poten- 
tial level of approximately 4.5 kv, again the point of intersection 
with the 45-deg line. In the case of the printer's gelatin roller, if 
film, charged either positively or negatively, contacts this roller, it 
will be brought to a level of about 1 kv. This occurs in a very few 
roller passages because of the steepness of the curve. Numerous 
types of curves are found with different roller-film combinations and, 
in many cases, quite different data are found on the support side from 
those on the emulsion side. 


In order to test the effect of rubbing film with a cleaning pad, a 2- 
in.-diameter roller was covered with the test material, and the film 
passed over the material on the roller with a 60-deg wrap, the roller 
being held stationary. The same technique of measurement as de- 
scribed above was used. 

It is found that ordinary velvet produces very little electrical charg- 
ing when rubbed against either the emulsion or the support side of 
Eastman Plus-X Panchromatic Negative (processed) Film. The 
curves lie very close to the 45-deg line (Fig. 4), showing that this film 
may pass across the velvet at any potential and its potential level will 
not be altered. This may, therefore, be termed a "neutral" combina- 

Mouton fur, on the other hand, alters considerably the potential 
level of processed motion picture negative film (see Fig. 5). In con- 
tact with the emulsion side, mouton fur discharges the film when it is 
charged to either positive or negative values, with the exception of 
positive potentials under 2 kv. With this exception, the film will al- 
ways leave the fur at a lower potential than it possessed upon reach- 
ing the fur. Low positive potentials will be increased but will not 
exceed a 2-kv level. This corresponds to the point of intersection 
with the 45-deg line. The mouton fur will discharge the support side 
very rapidly when charged positively. Between +4.5 kv and zero, 
the polarity is reversed by the fur, and for all approaching negative 
potentials, still higher negative values result. Successive passages 




+ 12 

+ 10 - 
+ 8 
+ 6 


+ 12 

+ 10 
+ 8 
+ 6 
+ 4 
+ 2 

- 4 

10 -8 -6 -4 -2 +2 +4 +6 

Leaving Potential V L 

-2 +2 +4 
vino. Potential V L 

Fig. 4. Electrification curves of dry Fig. 5. Electrification curves of 
velvet and processed motion picture mouton fur and processed motion pic- 
negative film. ture negative film. 

+ 12 

+ 8 
+ 6 

+ 12 

+ 10 
+ 8 
+ 6 

+ 2 


10 -8 -6 -4 -2 +2 +4 -t-6 

Leaving Potential V L 

Fig. 6. Electrification curves of Fig. 7. Electrification curves of velvet 
velvet plus petroleum ether and proc- plus Skelly Light Solvent and proc- 
essed motion picture negative film. essed motion picture negative film. 

will, therefore, build up very high potentials, since the curve does not 
intersect the 45-deg line, at least up to 6 kv, the limit used here. 
The point corresponding to V r = is of significance since it predicts 
the resulting potential with uncharged film, i.e., at zero level. A low 




positive charge equivalent to 0.5 kv will be imparted to the emulsion 
side, and a negative charge equivalent to 4 kv to the support side. 

If the velvet is wetted with petroleum ether and rubbed against the 
emulsion side of processed motion picture negative film, it will de- 
crease the potential of negatively charged film but will raise the po- 
tential of positively charged film (see Fig. 6). On the support side, 
the reverse is true, viz., positive potentials are reduced, and negative 
potentials increased. These may be termed "positive" and "nega- 
tive" combinations, respectively, since on the emulsion side, poten- 
tials move in the direction of positive values, and on the support in 
the negative direction, as shown by the arrows in Fig. 6. 

+ 12 

+ 10 
+ 8 

+ 12 

+ 8 
+ 6 
+ 4 
+ 2 

-6 -4 -2 O +2 +4 +6 +8 +10 KV 
Leaving Potential V L 

Fig. 8. Electrification curve of velvet 
plus carbon tetrachloride and the emul- 
sion side of processed motion picture 
negative film. 

-IO -8 -6 -4 -2 0,+2 +4 +6 +8 +IOKV 
Leaving Potential V L 

Fig. 9. Electrification curve of velvet 
plus carbon tetrachloride and the sup- 
port side of processed motion picture 
negative film. 

Velvet wetted with Skelly Light Solvent (a petroleum ether prod- 
uct manufactured by the Skelly Oil Co.) gives practically the same re- 
sults on both the emulsion and support sides as described above for 
velvet and petroleum ether (see Fig. 7). If velvet saturated with 
carbon tetrachloride rubs the emulsion side, it will maintain the poten- 
tial of the film at a constant value of +4.5 kv, regardless of the initial 
magnitude or polarity of the film potential. This is shown by the 
vertical lines in Fig. 8. It might be termed a "positive regulator." 

A similar phenomenon occurs when the support side is rubbed with 
velvet saturated with carbon tetrachloride (see Fig. 9), The regu- 


lated potential is also positive but has a much lower value, +0.4 kv. 
This represents the closest to the ideal found in any of the combina- 
tions tested, in that nearly complete de-electrification of the film is ac- 
complished. If film can be kept at low potentials of this order of mag- 
nitude, there should be little tendency for it to collect or hold dirt be- 
cause of electrostatic charges. 


The electrification behavior of film, when passed over a roller or 
rubbed with a cleaning pad, can be satisfactorily evaluated by bring- 
ing the film to given potentials, either positive or negative, and noting 
the change in the potential level after passing the test material. Most 
rollers of the more common materials which are suitable for use with 
motion picture films add charge to film rather than dissipate any ex- 
isting charge. 

Dry velvet does not appreciably change the potential of processed 
Eastman Plus-X Negative Film when rubbing either the emulsion or 
the support side. If velvet can be kept wetted with carbon tetra- 
chloride, it will hold this film at about +4.5 kv when rubbed against 
the emulsion side, or it will almost completely discharge the film 
when rubbed against the support side. 


The author is greatly indebted to R. Hubbard and T. Whitmore, for their as- 
sistance in this work, and to Dr. J. H. Webb, under whose direction this work was 
carried out. 


1. R. Gunn, "Principles of a new portable electrometer," Phys. Rev., vol. 40, pp. 

307-312, Apr. 1932. 

2. R. C. Waddel, "An electric field-meter for use on airplanes," Rev. Sci. Instr., vol. 

19, pp. 31-35, Jan. 1948. 

3. Ibid., see the bibliography. 

Variable-Area Sound Track Require 
ments for Reduction Printing 
Onto Kodachrome 



SUMMARY: This paper presents a plan for establishing the processing con- 
trol of variable-area sound tracks printed on Kodachrome. Data are pre- 
sented for two methods starting with the 35-mm master or dubbing print 
and following through the intermediate steps to the final Kodachrome 

THE INCREASED ACTIVITY in the Kodachrome field has made it 
necessary to establish commercial processing tolerances. The 
processing control of variable-area tracks in general has been success- 
fully established by the cross modulation method, and this same 
method was adopted to establish the processing tolerances for vari- 
able-area tracks on Kodachrome. There have been published a 
number of articles 1 " 4 on the processing control of variable-area sound 
tracks. All of these deal with the fundamental problem of controlling 
the image spread in the final print so that its average transmission 
will be constant regardless of the amplitude or frequency recorded on 
the track, provided that noise reduction is not considered. In the 
negative print process for black-and-white tracks this becomes a 
simple process of producing enough image spread in the negative to 
balance or cancel out the image spread in the print. It is understood, 
of course, that the print density is maintained high enough to give 
a good output level and signal-to-noise ratio. Kodachrome is a 
reversal process and this introduces other problems. The Koda- 
chrome sound track must be printed from a positive rather than a 
negative and the finished sound track is a silver sulfide rather than a 
metallic silver track. The exposed Kodachrome duplicates are 
developed in a normal black-and-white developer and the exposed 
silver bromide area is converted to metallic silver. The unexposed 
silver bromide is unaffected during this process. The sound track is 
then treated in a sulfide solution which converts the unexposed silver 

A CONTRIBUTION: Submitted June 8, 1949. 


46 ROBERT V. McKiE July 

bromide to silver sulfide. In the final processing step the metallic 
silver is removed and the portion of the Kodachrome track exposed in 
the printer becomes the transparent area of the Kodachrome print. 
The sound track remains as a positive image of silver sulfide. To pre- 
vent a serious loss of level, there must be sufficient exposure during the 
printing operation to maintain the clear area portion of the completed 
Kodachrome print relatively transparent. As shown in the following 
data this area becomes the controlling density. It is also necessary 
to control the image spread of the black-and-white printing master so 
as to cancel the image spread resulting from the Kodachrome printing 








Fig. 1. Methods for making Kodachrome sound tracks. 

A series of tests was planned to determine a practical method of 
establishing the processing control of printing variable-area tracks on 
Kodachrome. The following data are presented as examples of this 
control. Standard printers and developers were used in making this 
series of tests. The negatives and prints were processed according 
to the routine practice of the commercial laboratories printing Koda- 
chrome duplicates. 


Figure 1 shows the two methods most generally used for producing 
Kodachrome variable-area sound tracks by optical reduction printing. 
These methods will be discussed in the order shown. 


Method A 

A 35-mm cross modulation test negative was recorded at 80% 
amplitude. This test consisted of: (1) 400 cycles for reference level; 
(2) 4000 cycles for measuring high-frequency loss; and (3) 4000 cycles 
modulated in amplitude at a 400-cycle rate for cross modulation meas- 

Using Eastman fine grain sound recording film, Type 1372, the 
cross modulation test was exposed with a 3-mm 597 filter for a density 
of 2.70. This negative density value was determined from previous 
cross modulation tests for black-and-white printing. The negative 
was then developed in a high contrast variable-area negative 
developer at a gamma of 3.50. Previous tests had indicated that no 
advantages could be gained by varying the density of the original 

From the 35-mm negative a family of contact prints was exposed 
with unfiltered light onto Eastman fine grain release positive, Type 
1302. The black-and-white prints were developed in a print type 
developer at a normal release print gamma of 2.50. Print densities 
ranging from 1.28 to 2.25 were obtained. 

Figure 2 shows the cross modulation curve of the black-and-white 
prints that were used for making the Kodachrome duplicates. A 
normal release print at balance density (1.28), a balance density 
being the maximum cancellation point, and four other prints with 
increasing degrees of image spread as indicated by this cross modula- 
tion curve, were selected for making the Kodachrome prints. A lighter 
than balance density offered no advantages for Kodachrome printing. 

A family of black-and-white prints developed in a variable-area 
high contrast negative developer indicated that extremely high black- 
and-white positive densities on the order of 3.5 would be required for 
Kodachrome printing. As these high densities were often difficult 
to obtain and control under existing commercial printing conditions, 
only a print type developer was used for the final tests. 

Families of Kodachrome duplicates were then made by optical 
reduction printing from these 35-mm black-and-white prints. The 
Kodachrome film was processed at the Eastman Kodak Cine Proc- 
essing Plant in Hollywood in the normal manner for sound duplicates. 

The densities of the negatives, black-and-white positives, and Koda- 
chrome prints were measured with the Western Electric RA1100B 
densitometer using the visual filter. The Kodachrome prints covered 
a clear-area density range from 0.55 to 0.90. The Kodachrome prints 
are designated by the clear-area density rather than the sound track 
density. Due to the characteristics of the duplicating film, we have 




i i 

r 7 7 7 7 7 

SO Nl 13A31 

* * * 

I I I 

'S s 

I I 

i i 7 7 

80 Nl 1 3A 3 1 


found that the clear-area density is the best index of the image spread 
present in the Kodachrome sound track and therefore the most accu- 
rate means of measuring density for control purposes. 

The cross modulation tests were run on an RCA 200 16-mm repro- 
ducer through a calibrated reproducing system. The reproducer was 
calibrated with the SMPE multi-frequency 16-mm test film Series 

Figure 3 shows cross modulation curves plotted against the black- 
and-white print density. 

It has been established by numerous tests that 30-db cancellation 
of the 400-cycle component in the cross modulation test is satisfactory 
for all types of material; therefore density tolerances have been 
established at this cancellation value. 

From these curves it is evident that satisfactory cancellation may 
be obtained from black-and-white sound tracks having a wide range 
of densities. However, volume level, as shown by the 400-cycle 
curve, and high-frequency attenuation, as shown by the 4000-cycle 
curve, must also be considered. Therefore, the Kodachrome prints 
which most closely satisfy all the conditions of volume output, high- 
frequency response and cancellation would be a clear area density of 
0.74 printed from a black-and-white positive having a density of 
1.85. A lighter Kodachrome print density would require a darker 
printing master for sufficient cancellation with a resulting loss in high 
frequencies, as shown by the 4000-cycle curve of the 0.55 print density. 
A darker Kodachrome print density would result in loss of level as 
shown by the 400-cycle curve of the 0.90 print density. 

Method B 

Using Eastman fine grain sound recording film, Type 1372, a 
direct positive cross modulation test was exposed with a 3-mm 597 
filter over a density range from 1.60 to 2.30. The direct positive was 
developed in a print-type developer at a normal release print gamma of 

Figure 4 shows the cross modulation curve of the EK 1372 direct 
positive that was used for printing a family of Kodachrome dupli- 

Kodachrome prints covering a density range from 0.59 to 0.90 
were made by optical reduction printing from the 35-mm direct posi- 
tives. The Kodachrome film was processed at the Eastman Kodak 
Cine Processing Plant in Hollywood in the normal manner for sound 




j> ,L ^ ,;, j, A j. 

"O W <T t2> CO T & * W V ^ CD O 

< ?i7 < r7????7T 

80 Nl 13A31 


CM * 

? r r 


T T T 

80 Nl T3A3T 


These prints were measured in the same manner as the Kodachrome 
duplicates made under Method A. 

Figure 5 shows the cross modulation curves plotted against the 
direct positive density. 

The Kodachrome print which most closely satisfies the conditions 
of volume output, high-frequency response and cancellation would be 
a clear area density of 0.71 printed from a direct positive having a 
density of 1.88. The cross modulation test processed under these 
conditions also indicates that a lighter Kodachrome print density 
would require a darker direct positive density for sufficient cancella- 
tion with a resulting loss in high frequencies as shown by the 4000- 
curve of the 0.59 print density. A darker Kodachrome print would 
result in a loss of level as shown by the 400-cycle curve of the 0.90 
Kodachrome print density. 

For those studios not equipped to make cross modulation measure- 
ments, the proper combination of black-and-white and Kodachrome 
print densities can be determined by listening tests. If Method A is to 
be used, a short section of sibilant dialog should be recorded and 
developed to a normal negative density. A series of black-and-white 
prints made from this negative and covering a wide density range 
can be used for printing a family of Kodachrome duplicates. The 
Kodachrome prints should then be run on a good reproducer to deter- 
mine which combination of negative and print density gives the best 
quality. Improper density combinations will cause the sibilants to be 
distorted or rough. Therefore, the print which is free of sibilant 
distortion, which has the best volume output and high-frequency 
response together with low surface noise, will determine the proper 
combination of black-and-white and Kodachrome print density to be 

When printing from a direct positive the same procedure should be 
followed. The direct positive sibilant tests should be exposed over a 
wide density range and developed at a normal release print gamma. 
A family of Kodachrome prints made from the direct positives can 
then be run on a good reproducer to determine which combination of 
direct positive density and Kodachrome print density gives the best 


From the above tests it is evident that satisfactory sound quality 
on variable area Kodachrome prints may be obtained by selecting that 
printing exposure which will produce a clear area density giving satis- 
factory volume level and high-frequency response and by using a 


black-and-white with sufficient image spread to cancel the image 
spread which will be produced in the Kodachrome printing operation. 
From these tests the following values were found to produce the 
best sound quality for 16-mm duplicates made by optical reduction 
printing : 

For Method A 

(1) Negative exposed for a density of 2.70 and developed in a high 
contrast negative developer. 

(2) 35-mm black-and-white print exposed for a density of 1.85 and 
developed in a print-type developer at a normal release print 

(3) Kodachrome prints exposed for a clear area density of 0.74. 

For Method B 

(1) 35-mm direct positive exposed for a density of 1.88 and developed 
in a print-type developer at a normal release print gamma. 

(2) Kodachrome print exposed for a clear area density of 0.71. 

Due to variations in printers and developers, it is impossible to give 
absolute densities for the black-and-white printing masters to be 
used for making Kodachrome sound tracks. Therefore, data in this 
paper will apply to only one particular set of printing and processing 
conditions and can serve merely as a guide in helping to establish den- 
sity tolerances for other printing or developing conditions that will be 


1. J. O. Baker and D. H. Robinson, "Modulated high-frequency recording as a 

means of determining conditions for optimal processing," Jour. SMPE, vol. 
30, pp. 3-17, Jan. 1938. 

2. A. C. Blaney and G. M. Best, "Latest developments in variable-area process- 

ing," Jour. SMPE, vol. 32, pp. 237-245, Mar. 1939. 

3. Dorothy O'Dea, "Comparison of variable-area sound recording films," Jour. 

SMPE, vol. 45, pp. 1-9, July 1945. 

4. G. L. Dimmick, "High-frequency response from variable-width records as af- 

fected by exposure and development," Jour. SMPE, vol. 17, pp. 766-777 
Nov. 1931. 

The Pressurized Ballistics Range 
At the Naval Ordnance Laboratory 



SUMMARY: A description is given of the ballistics range at the Naval Ord- 
nance Laboratory, White Oak, Md. Details of the 25 photographic stations 
with their electronic controls are included. 

4 BALLISTICS RANGE is a piece of equipment used for determining 
_L\. the characteristics of a missile in flight. In its operation it is 
very similar to a classic experiment done by Leland Stanford and 
Edward Muybridge in 1872. Stanford, a wealthy sportsman, was 
interested in finding out information about the various gaits of the 
horse. He engaged Muybridge to set up a group of 30 cameras in a 
row, in a special building about 50 ft long. The shutters of the 
cameras were controlled electrically by wires stretched transversely 
from the cameras to a white wall on the opposite side. A horse 
galloping by touched the wires, and a series of pictures of the action 
were thus obtained. A chronograph measured the time intervals 
between successive pictures. 

If we use high-speed flash photography for the cameras, and sub- 
stitute electronic methods for the shutter mechanism, we will have a 
modern ballistics range. 

Figures 1 and 2 show the physical layout of the pressurized range 
at the Naval Ordnance Laboratory. The range is located in a steel 
tube 3 ft in diameter and over 300 ft long. Pressures up to five 
atmospheres and down to one-hundredth of an atmosphere can be 
obtained inside the tube. A standard 20-mm gun located in one end 
shoots the projectiles down the length of the tube into an 8-ft long 
barrier of sand. Smaller caliber guns can also be used. Twenty-five 
photographic stations are located along the tube. 

Figure 3 is a close-up of one of the photographic stations. When 
the missile passes between the source of light at A and a photocell 
located in B, a chain of events is initiated which ultimately causes 
the micro-second spark source, C, to flash. A shadow of the missile 
is thus thrown on the vertical photographic plate, D, and also by 
reflection from a mirror, E, to the horizontal plate, F. A set of 

PRESENTED: October 13, 1949, at the SMPE Convention in Hollywood. 





accurately located grooves from which the exact position of the missile 
can be determined is also photographed. 

Figure 4 shows the components required for one photographic 
station. These components will be briefly described' in the following 

Light-Screen Source 

This is a type T12-1 lamp manufactured for this purpose by the 
General Electric Co. The filament is approximately 28 in. long and 
is equipped with a spring tension device which holds it taut and 
straight at all times. The lamp is enclosed in a housing containing 
a slot covered with Eastman ruby safeiight material. This reduces 
to a low value any fogging of the photographic plates caused by the 
light-screen source. 

Photocell Amplifier 

The photocell amplifier is made up of a photocell, a three-stage 
amplifier, a thyratron and a power supply. The photocell is of the 

Fig, 1. Pressurized range, as seen from the gun end, 








Fig. 2. Pressurized ballistics range. 

high vacuum type with a red-sensitive photosurface. Light from the 
light-screen source is focused on this surface by a plastic cylindrical 
lens, which also serves to reduce the effect on the photocell of ex- 
traneous light coming from other directions. A gain control applies 
selective negative feedback to the amplifier and makes it possible to 
adjust the gain to the size of missile to be photographed. Because 
of the conditions of vacuum and pressure under which the apparatus 
is to be used, the paper condensers are of the hermetically sealed 
type, and no electrolytic condensers are used at all. 

Fig. 3. Perspective view of one photographic station. 

A, light-screen source C, spark E, mirror 

B, photocell amplifier D, vertical plate F, horizontal plate 

56 L. P. GIESELER July 

Fig. 4. Components required for one photographic station: foreground, 

General Electric Co. Type T12-1 Lamp and microsecond spark; background 

(from left to right), photocell amplifier, delay and trigger unit, high-voltage 
power supply and mam discharge condenser. 

With 70 volts applied to the light-screen source, the photocell cur- 
rent will be constant at approximately 0.5 microamperes, and no 
signal will be transmitted through the input condenser to the ampli- 
fier. A rapid change in light intensity such as is caused by a missile 
passing between the light-screen source and the photocell will orig- 
inate a signal which ultimately fires the thyratron. An output pulse 
of about 50 v magnitude will result, and this value is independent of 
the magnitude of the initial optical signal. 

Delay and Trigger Circuit 

This unit accomplishes the dual purpose of introducing an adjust- 
able delay in the sequence of events, and producing a 12,000-v trigger 
pulse which initiates the spark discharge. It consists of a 6J6 tube 
used in a "one-shot" multi-vibrator circuit, a 2D21 "booster" thyra- 
tron, a 3C45 output thyratron, and an output step-up transformer. 
The output of the first tube consists of a single negative rectangular 
wave whose duration can be varied from 70 to approximately 10,000 
/isec. The coupling network to the 2D21 thyratron produces a 
differentiating action, which changes the above signal to an initial 
negative pulse, followed after an adjustable interval by a positive one. 
The positive pulse fires the thyratron, which in turn fires the 3C45 
thyratron. A 0.5.-?/if condenser, which is initially charged to 1200 
v, discharges through both thyratron and the primary of the output 
transformer, producing an output of approximately 12,000 v. 

Spark and Discharge Circuits 

The output pulse of the delay unit is applied between a trigger 
electrode and one of the main electrodes of the spark. The ions 




Fig. 5. Electronic chronograph; seven standard counters and one 
test counter are available; the oscilloscope is used for trouble shooting. 


^BP^B^^Wl^^KP^^pP- ; ^B'^lBI|:^SK^::: 

Fig. 6. Close-up of four of the counters; an interval of approximately 0.1 sec 
has just been measured by the counters connected in parallel. 








formed cause the breakdown of the main gap, thereby discharging a 
0.4-jif main discharge condenser, and producing a short brilliant flash 
of small diameter. The effective duration is approximately 0.5 
/isec, and the intensity is sufficient to give pictures of good contrast at 
a distance of 72 in. At the instant that the spark flashes, a pulse is 
produced which actuates the chronograph. 

Further Considerations 

Lantern slide plates are used rather than film or paper to eliminate 
errors in missile position that might be caused by shrinkage. They 
may be either 11 X 14 in. or 14 X 17 in. in size. The large size is 
convenient for showing a large part of the shock wave and the turbu- 
lent wake. The developing is done in 80-gal stainless-steel tanks that 
will accommodate up to 48 plates at one time. Figure 5 is a photo 
graph of the electronic chronograph used to measure the time required 
for the missile to pass from a reference station to any other station. 
Intervals up to 1 sec may be obtained to a measuring accuracy of 
0.0000001 sec. The absolute accuracy is somewhat less, being deter- 
mined by a quartz-crystal oscillator which drives all seven counters. 
Figure 6 shows how four of the counters look after they have measured 
an interval of approximately 0.1 sec produced by the test counter. 
The four determinations of the interval read 0.0999975, 0.0999973, 
0.0999974, and 0.0999975 sec. Figures 7-9 are actual photographs 
of a 30-caliber bullet at various pressures. 

To explain the value of the pressurized feature of the range, it is 
necessary to discuss briefly some aerodynamic considerations. Most 
ranges and wind tunnels obtain information on small models rather 
than on the full-scale missile. For subsonic work, it is essential that 
the Reynolds numbers be the same for the two cases. The Reynolds 
number is equal to dVL/u, where d is the density, V the velocity, L 
a dimension on the model and u is the viscosity. For testing models 
in a supersonic flow, the important constant is the Mach number, 
which is the ratio of the velocity of the missile to the velocity of sound. 
The Reynolds number is, however, also important. With the pres- 
surized range it will be possible to vary the density of the gas inside 
the tube and thus to study the aerodynamic characteristics of missiles 
at the same Mach number but different Reynolds numbers. This will 
lead to a better correlation between model and full-scale data. 

An Experimental 
Electronic Background 
TV Projection System 



SUMMARY: The system is an electronic version of the process screen now 
used in motion pictures and television. Two television cameras may be 
mixed without super-imposition. A contrasting white background screen in 
back of the foreground subject, which is always brighter than the subject, 
provides contrasting information. This information keys the two cameras 
on or off through an electronic switch. Details of the signal selection, 
modification of the keying signals and the use of delay lines are discussed. 

BEFORE PROCEEDING we shall review a few basic principles of 
television in order more fully to understand the electronic proc- 
ess. Television is basically a scanning system; that is, the picture 
is divided into lines, and the transmitted signal is formed from a dot 
portion of a line which dot portion moves progressively from one end 
of the line to the other, retraces quickly back, and starts on another 
line. It takes 525 lines to make up one complete television frame 
and there are 30 frames in one second. Figure 1, trace A, shows an 
example of the electric signal corresponding to one horizontal line. 
The polarity of the signal as shown is white in the upward direction 
and black in the downward direction. The signal at each end of the 
line is the pedestal or blanking pulse which is black and which occurs 
during the retrace time of the scanning beam. It blanks out the 
retrace lines so that they will not show. 

In order to superpose the television signals from a foreground sub- 
ject and a background without securing a double exposure effect, a 
switching system is used to switch the output from the foreground to 
the background and back, as the scanning spot crosses the desired 
boundaries. This switch has to operate at a very fast rate in the 
order of 1/10 jusec (microsecond) or quicker. To select the desired 
boundaries a switching signal is used, and to derive the switching 
signal a contrasting signal is needed from the studio camera which is 
taking the foreground subject. The contrasting signal can be white 
or black; we prefer a white backdrop with the foreground subject 

PRESENTED: February 16, 1950, at the Pacific Coast Section Meeting in Los An- 
geles; and on April 25, 1950, at the SMPTE Convention in Chicago. 




performing in front of it. No part of the subject can contain any 
portion appearing as white as the backdrop, otherwise wrong switch- 
ing will occur in that region. 

The light intensity radiated from the white backdrop is 150 foot- 
Lamberts. This is about the right amount for the type 5820 camera 
tubes. The stop opening is adjusted for the saturation point, that is, 

*HIT -j j I __r r: 



BLACK-I 1 I 1 

u u 

Figure 1. 


Fig. 2. Block diagram of the system. 

at about //16. Trace B in Fig. 1 shows the camera video signal with 
only the white backdrop present. Trace C shows the camera video 
signal with the white backdrop and with a foreground object between 
the backdrop and the camera. 

As seen in the third trace, the white backdrop provides the con- 
trasting information. To satisfy the need for a means of electroni- 




cally isolating the white backdrop signal from the foreground object, 
an amplitude selection method is used, that is, a portion of the video 
signal is selected by the grid cutoff characteristics of electron tubes. 
The upper trace, C, shows the video signal at the grid of a tube ; the 
dotted line marked "clip level" shows the grid cutoff point. The 
portion of the signal between this and the clear white level yields the 
resultant plate current, in the lower trace, D, which is in a form that 
could be used directly as a switching signal to switch between the fore- 
ground camera and the background information that it is desired to 
mix with the foreground subject. 

Figure 2 shows a block diagram of the system. The foreground 
camera is in the upper left portion of the diagram. The white back- 
drop is in front of this camera and between the two is the foreground 

T3 T4 


Tl T2 

Figure 3. 

subject. The video information from this camera then goes to a 
distribution amplifier which bridges the line and sends the video 
signal to the amplitude selection amplifier. This video signal also 
goes through a separate branch to form the pictorial signal when the 
foreground subject is switched in. Delay lines are used in both 
branches so that the times of transition in the foreground and switch- 
ing signals match exactly at the final switching point. The output of 
the amplitude selection amplifier goes through a pulse narrowing 
system which will be referred to later. The switching signal from the 
latter system then operates a diode switch which has connected to one 
side of it the foreground video signal from the delay line. The back- 
ground video signal from a camera or from a motion picture or slide 
projector is connected through an identical delay line to the other side 




of the diode switch. The switching diodes are connected in opposite 
polarity so that, when one is on, the other is off. The common output 
of the diode switch is bridged by a distribution amplifier which changes 
the output impedance to 72 ohms and sends the composite mixture 
of foreground and background picture to the studio switching system. 
Figure 3 shows in trace A an expanded view of a horizontal line 
from the foreground camera. The camera output at time T-l is at 
white level and it is going to black level which is the foreground sub- 
ject. Prior to time T-l this camera was looking at the white backdrop 
and its output did not appear in the composite mixture, but was re- 
placed by the background information. Television cameras do not 
have infinite detail so the transition from white to black requires a 
finite time. The speed of transition varies from 0.06 //sec for 600-line 
definition to 0.12 /isec for 300-line definition. It is this limitation 


Figure 4. 

that makes necessary a pulse narrowing device. Suppose the ampli- 
tude selector selects the signal at time T-l and develops a switching 
signal at the same time if the background information at this time 
is dark as shown in trace B it will, unfortunately, be turned off too 
early. Trace C shows the resultant video signal when this occurs. 
As the figure shows, a short white pip from the backdrop around the 
foreground subject will be obtained. The net result of successive 
horizontal lines is to produce a white ring around the foreground 
subject which is called halo. The same thing happens in reverse 
order when the switch turns the foreground subject off and the back- 
ground subject on, that is, during the time from T-3 to T-4. 

In order to correct for this limitation it is necessary to delay the 
switching signal from T-l to T-2 and in reverse order advance the 
switching signal from T-4 to T-3, as shown in trace D of Fig. 3. Of 
course, there are no advance lines being manufactured at this time 




Fig. 5A. The background scene. 

Fig. 5B. The foreground actor, with white backdrop. 




and a different scheme is used involving the narrowing of the switching 
signal by means of a delay line. Figure 4 shows in the upper portion 
a delay line connected to the output of the amplitude selection ampli- 
fier. A 300-ohm line is used with the far end left open. The trace 
at the bottom of Fig. 4 shows the switching wave form as it is modified 
by the delay line. At time T-l the voltage is one-half of normal 
because the driving impedance is in parallel with the characteristic 
impedance of the delay line. When the switching voltage returns 
J'fo Msec later from being reflected at the open end of the delay line 
it adds to the original signal at time T-2. At T-3 the switching 
voltage from the amplitude selector goes to zero, but the reflected 
signal from the open end of the delay line continues for an additional 
Ko Msec. The modified switching signal is clipped by another ampli- 
tude selector which selects the narrowed portion of the pulse giving a 
switching signal that is Jf o Msec narrower. In the process of narrow- 
ing the switching pulse, an additional delay of 0.05 jusec has been in- 
curred. This delay is compensated for by the video delay lines before 
the video signal arrives at the switcher. The timing has to be held to 
0.01-jusec accuracy in order to prevent any trace of halo. So much 
for the electrical features of the electronic background projection 

Fig. 5C. The composite picture. 


Some of the characteristics of the composite mixture are of interest. 
For instance, the relation of the foreground picture to the background 
picture during the process of dolly shots and panning presents some 
problems and also offers some possibilities for unusual effects. The 
dolly shots are very realistic when the background picture is of an 
outdoor scene. The background picture remains stationary if it is a 
still, while the foreground picture is changing, that is, if the camera 
starts to dolly at a distance the actor first appears small, and then as 
the camera moves in the actor fills up more and more of the screen. 
At the same time, however, the background does not change. This 
gives the illusion that the background is at a considerable distance. 
If, on the other hand, the foreground camera is panning, some strange 
things appear to happen. The background remains stationary while 
the actor moves left or right as the foreground camera is panned. 
This gives the appearance that the actor has been moved by an 
invisible hand; there is no body action to indicate movement. If 
this effect is not desired the camera should be locked in azimuth and 
elevation. In normal shots this is not an insurmountable difficulty, 
and ways and means are being developed to eliminate it. 

The nature of the system depends upon amplitude selection and 
this, therefore, places a limitation on the type of clothing that can be 
worn by the foreground performers and the type of lighting to be 
used in the working area. No white clothing should be worn since 
none of the high lights in the foreground subject should be greater 
than the reflection from the illuminated backdrop. The desired type 
of clothing would be in the gray region. Black clothing does not give 
too satisfactory a picture when image orthicon tubes are used in the 
camera. Large white areas on the image orthicon target cause some 
secondary electrons to be emitted into the darker areas of the target 
and causes the black areas to become lighter than they should. 
While normal type of makeup can be used in most cases, it has been 
found in some instances that a heavier type of makeup helps to subdue 
spectral reflections. Makeup on the hands has to be used for the 
same reason. The area in which the foreground subject performs 
needs to be lighted somewhat differently from the normal shots. A 
constant intensity of lighting is needed in the working area and this 
means that the light sources should be located farther away than 
usual. We have been using the long slim-line floor fluorescent lamps. 

Figure 5A shows the background scene, and Fig. 5B the foreground 
actor with the white backdrop. In Fig. 5C is illustrated the com- 
posite picture obtained from the processed signal. 

Effects of Incorrect Color Temperature 
On Motion Picture Production 


SUMMARY: Past efforts to systematize control of film production (and 
especially color) have been partially defeated by inability to detect variations 
of color temperature of daylight and artificial light sources. Effects of such 
variations on tone or color of makeup, costumes and sets are cited. Steps 
necessary to complete control, and the part which color temperature of illumi- 
nants plays in each are indicated. A new instrument is described which 
makes the practical determination of color temperature a simple step. A 
more accurate method is proposed for the description of illuminants, color 
film and filters. 


rriHE TIME HAS PASSED when it is necessary to explain to a body of en- 
_L gineers what we mean by the "color temperature" of an illuminant. 
The concept of a black-body radiator and its Kelvin temperature is 
generally understood by illuminating engineers, studio technicians 
and workers in the field o color photography. 

If the illuminants in general use were true black-body radiators, and 
if their color temperature were a factor of high constancy, there would 
be no problem in the application of such light sources to color photog- 
raphy and cinematography. Unfortunately, we are becoming 
increasingly aware that practical light sources represent only ap- 
proximations, and often poor ones, of black-body radiators. It has, 
of course, long been realized that practical light sources vary con- 
stantly in their spectral distribution of energy, with serious results 
upon the exposure balance of color materials. In the earliest litera- 
ture of color photography, there are frequent references to the insta- 
bility of filter factors caused by the fluctuating color balance of light 
sources, and especially daylight. On the whole, in reading these early 
statements, one is struck by the fact that the subject was less compli- 
cated and easier to understand before the concept of equivalent color 
temperature was introduced. 

In the following paper, we shall sometimes use the term "color 
temperature" for lack of a generally accepted term which would be 
more descriptive. We shall also, however, set forth the first results 
of a series of studies in the measurement of the color balance of practi- 
cal illuminants, and make certain proposals concerning a more simple 

PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. 



and informative method of designating this property of a light 

Until fairly recently, color temperature might be said to have been 
the "hidden factor" in color film production and color photography. 
We knew about it, but because there was no available means of taking 
it into account, there was a natural tendency to ignore it. The color 
balance of a scene sometimes turned out correctly, and often did not, 
and when it did not the blame was usually placed on the manufacturer 
or the laboratory. The manufacturer, harassed by the difficulties of 
making color film with reasonably constant over-all speed, went on 
the simplifying assumption that the color temperature problem could 
be solved by the user of color materials. The processing laboratory, 
busy with the complexities of maintaining unstable baths at constant 
energy, likewise assumed that the color temperature of the taking 
illuminant ignored. 

This simplification was undoubtedly justified as a practical measure 
during the period in question. Today, however, it is neither justified 
nor necessary. Now, that incident light measurement has been 
generally accepted as a means of exposure determination, the ASA 
Exposure Index 1 takes on real meaning, and the recent announcement 
of Eastman Kodak that professional color film materials will be pack- 
aged with a slip bearing the individual exposure index of that particu- 
lar batch will mean that exposure determination will take on an 
increased degree of precision. 

Improvements in laboratory procedure and the use of color densi- 
tometry now make it possible to hold the color balance of monopack 
or integral tripack materials to tolerances comparable with those in 
handling separation negatives on black-and-white film. Therefore, 
it would be fair to assume that results should now be perfect, but, 
as every practical worker knows, they are not. There must be 
another serious factor at work, affecting color balance and this fac- 
tor, of course, is the color balance, or temperature, of the taking light. 

We believe, therefore, that the time has come to bring color tem- 
perature out into the open, to recognize it, and to try to solve the 
problems which it creates, just as these other problems have been 
solved. In addition to intensity of illumination, brightness range 
and contrast, we must now measure the color of the illuminant. 

Even in black-and-white work, this hidden factor of color tempera- 
ture has had an important effect on tone reproduction of different hues. 
Every cameraman has had the experience of being unable to duplicate 
the results of an earlier test, even when film bearing the same emulsion 
number was used. All the conditions seemed the same, but the fact 


that he may have been given old lamps for the test and new lamps on 
the set, or may have used "nets" instead of dimmers, may have been 
enough to shift the spectral distribution of the illuminant, and with it 
the spectral sensitivity of the film. 

Even now, this question would be purely academic if no instru- 
ments were available for the measurement of color temperature. 
Such instruments are available, however, and their use will undoubt- 
edly become an integral part of professional procedure. The pro- 
fessional worker knows that if certain factors are to be measured and 
controlled, then all factors should be similarly measured and con- 
trolled, or the final result may be worse rather than better. 

Nor should it be felt that this complete application of measurement 
restricts the freedom of the artist. The cameraman, as artist, will 
still be free to deviate from the norm in any direction which he desires, 
with the added benefit of the assurance that the results will be what 
he wishes. 

Before dealing further with the applications of color temperature 
measurement in motion picture production, we shall describe briefly 
the problems of designing an instrument for this measurement, and 
give one solution of the problem. 

Awareness of the effects of incorrect color temperature on color 
balance is not new. When Wall wrote his Practical Color Photog- 
raphy 2 in 1922, he said: "Theoretically, one ought to determine the 
filter ratios before each exposure, as the color composition of daylight 
varies considerably, being much richer in red and green in sunlight 
than in shadow or in cloudy weather. But if the filter ratios have been 
determined, one may ignore this factor, at any rate at first." Consid- 
erably earlier references may be found in von Hubl, Koenig, and oth- 

Despite this awareness, surprisingly few efforts have been made in 
the past to provide the photographer with a means of measurement 
and control. Two visual instruments have been placed on the market. 
The first was based on the visual match of a yellow filter and an 
additive mixture from red and green filters. The second relies on a 
"dichroic" filter which appears pink under one type of illuminant 
and bluish under another. Such meters have undoubtedly been of 
considerable assistance when used correctly, but are unavoidably 
affected by the adaption and condition of the observer's vision. 

For this reason, it was felt worth while to develop an instrument 
which would not involve judgment, or any subjective factors, despite 
the many design problems involved. 

The term "color temperature" has three meanings at the present 


time : a particular spectral distribution of energy, a visual appearance 
which matches a particular spectral distribution, or (in photographic 
literature only) a certain ratio of photicities in three broad zones of 
the spectrum. 

The instrument in question has been designed with a full awareness 
of this triple nature of color temperature, and to a high degree, the 
instrument is suitable for the measurement of all three. 

If radiation is close to the true spectral distribution of black-body 
radiation, then the ratio of the measurement at any two points in 
the spectrum will be characteristic of one color temperature only, and 
if the instrument is properly calibrated it will be possible to measure 
the color temperature of true black-body radiation accurately over the 
desired range. This can be done with an accuracy well within the 
range of least perceptible differences. 

Any two points in the spectrum might serve, but for maximum 
precision it is desirable to have the distance between the two points 
great enough so that filters with no overlap may be selected. 

In the case of visual color temperatures, the instrument becomes 
in effect a photoelectric colorimeter, and the visual sensitivity curves 
used in standard colorimetry must be taken into account. Any two 
of these might be used, but since it has already been decided that 
there shall be no overlap, the best choice would seem to be the blue 
and the major portion of the red. The fact that the secondary maxi- 
mum of the red sensitivity curve (in the blue) is not present in the red 
filter becomes without significance in this case, since the secondary 
maximum would alter the blue-red ratio by a constant factor, and 
this cancels out in the calibration of the instrument. 

Thus, in the case of light sources with continuous spectra, the in- 
strument will give the color temperature of the black-body radiation 
for which the illuminant in question is a visual match. 

In the third case, that of photographic color balance, the problem 
is simple and straightforward. We are interested in measuring the 
amount of red and blue, and sometimes green, in three relatively broad 
zones with well defined maxima. While different color processes vary 
somewhat in the boundaries of these zones and in the precise location 
of the maxima, the similarities are far greater than the differences, 
and the agreement is close enough to make it entirely feasible to 
select reference points for measurement which will be valid for all 
present processes with precision well within the permissible toler- 

The same red and blue filters which were selected to meet the second 
condition also serve for the third purpose, that of measuring photo- 




graphic color balance. When departure from the correct amount of 
green is suspected, this can be checked by the use of the third filter, 
green, with which the new professional model of the instrument is 
equipped. If the amount of green is not that for which the red-blue 
ratio would call, in accordance with black-body standards, the meter 
will at once indicate it, and corrective measures may be applied. 

The curves shown in Fig. 1 represent the effective sensitivities 
resulting from the combined effect of filters and barrier type cell. 

How this is accomplished will immediately be clear from the follow- 
ing description. The body of the instrument consists of a circular 
housing, about 4 in. in diameter, with a pistol-type grip on the bottom 
for convenience in aiming the instrument at a light source. The body 
has a gray crackle finish. On the front of the circular housing is the 
opening through which light enters the instrument, surrounded by a 
knurled ring which is coupled to the diaphragm over the light cell. 
This diaphragm consists of two overlapping sectors with serrated 

Fig. 1. Comparison of the standard 
blue and red visual sensitivity curves 
(dotted curve and dashed curve) with 
the sensitivity of barrier cell plus blue 
and red filters in a commercial color 
temperature meter. 

edges. Behind this diaphragm is a red filter, through which the light 
passes before reaching the barrier-type cell. 

On the back side of the circular housing, set at an angle for easy 
visibility, is the needle and scale of the milliammeter. On this scale 
is a red mark. The user points the instrument at a light source and 
adjusts the diaphragm ring until the needle coincides with the red 
mark. All that remains is to pull the trigger set in the front of the 
pistol grip. This replaces the red filter with a blue filter, and the scale 
reading then gives directly the color temperature of the light source, 
or the deviation from a standard temperature and the correction filter 
to use to bring the illuminant to that standard, depending on the 
scale in which the instrument has been calibrated. 

One accessory to the instrument should be mentioned. This 
is a frosted glass hemisphere, completely neutral in color, which may 
be fitted over the front of the meter. This is for use when the prevail- 


ing illumination is not all of one color, and it is desired to obtain a 
practical average value. 

When measurement of the green is desired as well, for illuminants 
which are not a good approximation of a black-body radiator, it is a 
simple matter to add a green filter and a second trigger position. 
This is now being done on a new professional model which will be 
ready shortly. The significance of this green measurement, and the 
relation of readings to the use of corrective filters, will become clearer 
in a later section of this paper, in the discussion of a proposal for an 
improved method of specifying the color balance of an illuminant. 
We will first consider some of the possible applications of color tem- 
perature measurement to actual studio procedure. 

When we speak of color temperature today in the motion picture 
industry, as already mentioned, we may mean any one of three quite 
different things: (1) We may mean true black-body radiation, or a 
close approximation thereof. (2) We may mean a light source which 
is a visual match for such black-body radiation, even though its 
spectral distribution is vastly different (and such a light source corre- 
sponds to the meaning of the adopted American nomenclature) . (3) 
In purely photographic and cinematographic terminology, by color 
temperature we may mean merely the proportions of red, green and 
blue radiation in a light source, which balance the sensitivities of a 
particular type of color film a light which may be neither black- 
body radiation nor a visual match for such radiation. 

This third usage of the term is not met with in the scientific litera- 
ture, properly speaking, but is the common usage in motion picture 
practice, and as such is worthy of a little more consideration than 
has as yet been given to this subject. In motion pictures, and in 
color photography generally, we are not concerned with the con- 
formity of a light source to a particular black-body distribution of 
radiation. What we are interested in is a continuous spectrum in 
the three broad bands or zones of the spectrum which will be re- 
corded on the film, and it is desirable that the general curve of spectral 
energy distribution be reasonably smooth, so that it will not show any 
unpleasant surprises in connection with colors having narrow absorp- 
tion or reflectance bands. With this qualification it is only necessary 
that the energy which falls in the band of the blue filter or sensitiza- 
tion, the green filter or sensitization, and the red filter or sensitization 
have a proper ratio to the total of the three or to each other, since this 
will ensure color balance. Actually, it would be more accurate and 
correct, and more easily intelligible, to the average nonspecialist in 
colorimetry if we were to use the term "color balance" of the illumi- 


nant rather than color temperature, since it is color balance that is 
actually meant. 

In setting up a color process, whether it be one which uses three 
separate negatives, like Technicolor, or an integral tripack coated on a 
single base, such as Kodachrome and Ansco Color, the maker balances 
the film for a particular color ratio in the taking light source. So far 
as we are aware, very little information has been released by Eastman 
Kodak and Ansco on the procedures involved in this and the standards 
used, and it might eventually become an important contribution to a 
better standardization and simplification of nomenclature in this field 
if the manufacturers were disposed to co-operate in setting up speci- 
fied procedures. It must not be forgotten that the illuminant under 
which color motion pictures are taken is not an end in itself but solely 
a means to an end, that end being a positive color film which the 
ultimate beholder will find agreeable and acceptably realistic. 

Color temperature, in motion picture production, is not an academic 
question. Given a particular color process and a particular batch of 
film, what is desired is to make a balanced set of negatives on that 
film, using whatever light may be necessary for that purpose regard- 
less of what this light may be called or how it may be classified. The 
basic conditions to be satisfied in this direction were laid down many 
years ago, when the practice of color photography was in an almost 
purely empirical state, in terms of what is known as "the first gray 
condition" and "the second gray condition." The first gray condition 
specified that a neutral gray, photographed through the three taking 
filters, shall produce equal densities in all three negatives, and the 
second gray condition specifies that equal densities in the three nega- 
tives shall produce a neutral gray, or a good approximation thereof, 
in the finished positive. Today, we should probably modify that a 
little bit, recognizing that in such processes as color development and 
some other means of forming colored images, equal silver densities 
do not necessarily yield equal dye densities. Therefore, the "first 
and second gray conditions" could probably be more accurately 
reworded to say that "a neutral gray object, photographed through 
the three filters, shall produce correctly balanced densities in the three 
negatives, and that correctly balanced densities in the three negatives 
shall be those densities which produce dye densities in the positive 
giving the best neutral gray of which the process is capable." 

The simple statement of the problem in this way holds the whole 
basic question of photographic color temperature, and helps to point 
up the fact that what we are interested in is not color temperature 
per se, but a certain specific and definite ratio of silver densities in the 


negative image and a certain ratio of dye densities in the finished 

Nothing will emphasize the absurdity of the present color tempera- 
ture nomenclature in relation to photography and cinematography 
more than the necessity of explaining the matter to someone previously 
quite ignorant of it. At the present time, one is first obliged to explain 
the concept of a black-body radiator, an abstraction not too easily 
grasped by the nontechnical mind, then the complex mathematics of 
determining the distribution of energy in the radiation from a black 
body at different temperatures, the character of light emitted at the 
different temperatures, and so on. After this somewhat lengthy be- 
ginning, it is then necessary to confuse completely the person to whom 
the explanation is being given by explaining that what we have just 
told him is color temperature, but that what we are talking about is not 
that but something quite different. Then, it is necessary to go into 
the explanation of equivalent color temperature and light sources 
which are a visual match for a particular color temperature, after 
which we must again explain that that is not what we mean either, 
but that color temperature in relation to photography actually means 
the balance of red, green and blue in certain important sections of the 
spectrum, such balance to be similar to the balance of those same 
zones in a black-body radiator. 

"Color balance of the illuminant" is probably too clear and simple 
a term to find favor as a new nomenclature, but there assuredly is a 
drastic need for a simple term which will make it clear that we are 
concerned with the balance of three specific zones and not with black- 
body radiation or a visual match for it. If, as we believe to be wise, 
a new term is sought, it would seem desirable at the same time to 
adopt a more rational system of numerical evaluation than that 
employed in the color temperature system. Judd 3 pointed out 14 
years ago that the use of the reciprocals of Kelvin temperatures would 
give rise to a scale in which the least perceptible difference in color 
temperature to the observer would remain more or less constant, 
whereas on the Kelvin scale it differs sharply from zone to zone. 
At 3200 K (degrees Kelvin) for example a difference of slightly less 
than 50 K is a perceptible difference, whereas at 6500 K, the least 
perceptible difference is in excess of 200 K. On the other hand, if 
these are expressed in terms of Micro-Reciprocal Degrees or Mireds 
(obtained by dividing the Kelvin temperatures into 1,000,000) we 
find that the least perceptible difference represents about the same 
number of Mireds over the entire portion of the scale in which we are 
interested. A temperature of 3200 K becomes 312 Mireds, and 6500 


K becomes 154 Mireds, and in both cases the least perceptible differ- 
ence would be approximately 5 Mireds. This least perceptible differ- 
ence is not a constant throughout the whole of the Kelvin scale, but 
as Judd 3 has pointed out, it is substantially constant from 1,800 to 
11,000 K, which fully covers the range in which we are interested. 

There would seem, therefore, to be every practical advantage 
in specifying light sources in terms of Mireds rather than degrees 
Kelvin, when we are discussing their visual appearance. However, 
the Mired, as a unit, still fails to convey directly the information in 
which we are interested for photographic purposes, and we shall pro- 
pose a further standard later in this paper, though we shall relate it to 
the Mired. 

One field which calls for definite investigation is the establishment 
of tolerances which will specify the permissible variation in color of the 
taking light. We know about what this should be when we are deal- 
ing with direct visual perception, but we do not know what difference 
in the color of the illuminant will produce a just perceptible difference 
in a color photograph taken by that illuminant. 

Presumably, since the color photograph is of lower saturation, the 
tolerance is somewhat greater. Presumably, also, the better the sub- 
tractive primaries the more critical the balance, the poorer the prim- 
aries, the greater the tolerance. We know, for example, that when a 
letterpress printer has difficulty in controlling the balance of a three- 
color job, he deliberately "grays" the process inks by contaminating 
them with each other, so that the balance will be less critical. 

We may assume that in an ideal process of color photography, the 
photographic tolerances would be the same as the visual tolerances, 
or about 5 Mireds. As to actual processes, we have little data. Dr. 
Spencer, in an investigation made in England before the war, found 
that a density variation in a Carbro separation positive of about 5% 
represented the permissible limit. This is roughly equivalent to 10 
Mireds illuminant color difference, or two visual steps. Eastman 
Kodak recommends a tolerance of about the same amount for Koda-. 

So, while we should not infer too much from these unrelated obser- 
vations, as good a guess as any at the present time would be that on 
current processes the photographic tolerance is about twice the visual 

However, it is not suggested that the use of this doubled tolerance 
in the exposure of color film would be good practice, for two reasons : 
first, because as processes improve, the photographic tolerance will 
approach the smaller visual tolerance, and second, because if we utilize 


the full tolerance of imbalance at the time of shooting, no tolerance 
is left for the manufacture or processing of the film. 

A practical and sound tolerance, then, which we believe should be 
recommended at this time, is =*= 5 Mireds. At tungsten temperatures, 
this represents about ==50 degrees K, and at daylight temperatures, 
200 degrees. 

Permissible variation in directions away from the black-body locus 
remains to be investigated. 

Before going farther, we shall consider the steps involved in complete 
color control in the studio, after which we shall take up a proposed 
means of systematizing such control. The major steps which affect 
the validity of color reproduction in a motion picture are the following: 

1. Selection of correct subject matter. 

2. Use of illuminants of correct color balance while shooting. 

3. Correct color balance of film sensitivities and filter transmis- 

4. Use of a camera objective which is reasonably nonselective. 

5. Balanced negative processing. 

6. Balanced printing of the positive. 

7. Balanced positive processing. 

8. Projection with light of uniform color, both in distribution and 
duration (and preferably as white as possible), and with a minimum of 
stray light reaching the screen. 

9. A screen which is reasonably nonselective. 
10. An observer with normal vision. 

With slight exceptions, these steps apply with equal force to all 
processes in use at the present time, and we shall briefly consider the 
part which the color balance of illuminants plays in several of these 

1 . Selection of Subject Matter. This involves the choice of fabrics 
for costumes, pigments for set decoration, cosmetics for makeup, 
and many other items. Some of this choice is done by visual color 
.matching, some by practical tests shot in advance. In the case of 
visual matching, it seems to us particularly unfortunate that there is a 
general tendency in the studios to do this under fluorescent lighting. 
As Nickerson, 4 Evans, and others have recently pointed out, the 
presence of strong blue and orange monochromatic bands in these 
illuminaires leads to considerable distortion of colors with fairly 
abrupt absorptions and reflectances. The use of properly filtered 
tungsten sources would lead to more consistent and reliable color 
matching and selection, particularly in the makeup department. 
Such sources should be checked frequently for proper color with a 


suitable meter. With standardized color matching sources of this 
type, makeup colors could be standardized, ending the present chaos. 
Different studios, using the same Technicolor process, employ sharply 
differing basic makeup colors. Which are correct? Which are bet- 
ter? It would seem that this might well be a matter for the attention 
of the Research Council of the Academy of Motion Picture Arts and 
Sciences, rather than the individual manufacturer of cosmetics. As 
regards camera tests of makeups, fabrics and the like, it goes without 
saying that the color of the illuminant should be rigorously controlled, 
so that it may be duplicated during production. 

2. Taking Illuminant. The causes of variation in the color of the 
prevailing illumination on the set or on location are so numerous that 
an entire paper could easily be devoted to them. In the studio, the 
type and age of lamps, the line voltage, silks which grow yellow, niters 
which fade, arcs which smoke these and a score of other factors 
make the color of set illumination problematical. In the open air, 
there are comparable variations due to meteorological and geographical 
conditions. We know that reasonable errors in balance can be cor- 
rected later, provided the error is fairly constant over the entire frame. 
Nothing can be done, however, if it is a single face, or a single portion 
of the set. Control by means of suitable instruments will reduce the 
need for laboratory correction to a minimum, and should virtually 
eliminate scenes which cannot be corrected. 

3. Film and Filter Balance. We know that there is some unavoid- 
able variation in manufacture, in the age of the film, and so on. How- 
ever, if we hold our tolerances closely on the color of the taking 
illuminant, the requirements of a particular emulsion number can be 
determined accurately and closely met. Naturally, we must be sure 
there is no serious image regression through too long storage after expo- 
sure and before development, since this affects balance very adversely. 

4. Nonselective Objective. This is a minor item, but yellowed balsam 
in an old lens can affect blue transmission perceptibly, and low- 
reflectance coatings which are all of the purple type can drop red and 
blue transmission as much as 6%. All blue or all yellow coatings 
give an even worse result. The remedy, of course, is to discard old 
and yellowed objectives, and to have low-reflectance coatings applied 
with a suitable mixture of purple and brown surfaces. 

5. Negative Processing. This has been so adequately dealt with in 
other papers before the Society that there would be no point in 
repetition here; however, better standardization of illuminant, film 
and filter relationships will obviously make it easier to obtain uniform 
laboratory results. 


6. Balanced Printing. This again calls for accurate control of the 
color of the illuminant (especially in the case of multi-layer materials) . 
Tolerances should be held as closely as on the set, which means photo- 
electric measurement. 

7. Positive Processing. This has also been covered. If previous 
steps have been held within desirable tolerances, this step should 
offer less difficulty than at present. 

8. Projection Light. The light reaching the screen should be 
reasonably white, uniform in color over the screen, and uniform in 
color when making a change-over. 

9. Projection Screen. Should be reasonably nonselective. Photo- 
electric control is useful to check the combined performance of light 
source and screen. 

10. Normal Observer. If the observer is color blind, there is nothing 
we can do about it. We can, however, be reasonably sure that the 
responsible personnel working with color have normal color vision, 
since the unsuspected presence of an individual with some form of 
color blindness can cause much waste and confusion. 

So much for the steps in production at which control may be exer- 
cised. We have purposely curtailed this section somewhat, because 
we feel that more importance attaches to a proposal which we shall 
make for a new system of measuring and specifying the color balance 
of illuminants instead of the Kelvin color temperature scale. After 
all, an adequate system of measurement and description is the essen- 
tial first step toward better control, so all of this is extremely pertinent 
to the general subject of the paper. 

Dissatisfaction with the term " color temperature" and all that it 
stands for is not precisely new, but it has recently become insistent. 
Nickerson 4 said, in a paper at the last meeting of the Society: 

"The specification of the color of sources in terms of color tempera- 
ture without an understanding of the limited meaning of the term 
has caused much confusion in color thinking as it concerns the illumi- 
nant . . . For any real understanding of color processes, whether 
visual or photographic, it is necessary to take into consideration the 
more exacting specification of spectral distribution. Thus, while 
illuminants in this report are often referred to in terms of the color- 
temperature scale, it should be remembered that it is not their color 
but only their spectral characteristics that will tell whether they are 
suitable for use with a given film, or to produce a specified result." 

Evans 5 says, in his invaluable book: "The usage of the term is 
exceedingly confusing and it appears inevitable that sooner or later 
a new terminology will appear." 




Moon 6 says, in the report of the Optical Society of America Com- 
mittee on Colorimetry : " ... the concept of color temperature . . . still 
serves as a rough engineering specification. However, one may 
expect it to decline gradually, to be replaced by the much more satis- 
factory spectrophotometric curve and by the colorimetric methods of 
Chapter XIII." 

To this, Jones and Condit 7 add the following comment in a recent 
paper before the Optical Society of America: "Our feelings are some- 
what stronger than those of Moon concerning the discontinuance of 
the usage of the term 'color temperature/ particularly in the field of 
photography. We should like to recast the last sentence of the above 
quotation to read as follows : 'However, we hope its use in the field 
of photography will decline rapidly or cease abruptly and be replaced 
by... etc." 

/ \ 







\ 1 

\ 1 
\ 1 




\ t 








Fig. 2. Two monochromatic lines, at 
470 and 573.1, visually match Illu- 
minant C but would photograph as 
pronouncedly blue, as shown by dotted 
trichromatic filter curves. 

Fig. 3. Two monochromatic lines, at 
490.4 and 610, visually match Illu- 
minant C but would photograph as red. 

Our own work in the field of illuminant color measurement has made 
us extremely conscious of the shortcomings of Kelvin temperature, 
which has led us to seek something more descriptive. 

The inadequacy of visual-match color-temperature as a standard for 
color photography can be best shown by taking a few extreme cases. 
In Fig. 2 is shown the spectral distribution of a light source which 
emits two monochromatic lines, as shown. Colorimetrically, this is a 
visual match for Illuminant C, the artificial daylight of colorimetry. 
Photographically, as shown by the three filter transmissions indicated 
in dotted lines, this illuminant would photograph as a bright blue. 
The source shown in Fig. 3 would photograph as bright red, yet 
visually it matches Illuminant C. Lastly, that shown in Fig. 4 also 
matches Illuminant C, and would scarcely record on normal color film 
at all. 




These are extreme cases, but it would be a mistake to assume that 
they are irrelevant. The rising popularity of lamps with strong 
monochromatic lines, even though these be superimposed on a con- 
tinuous spectrum, makes it necessary to stress as strongly as may be 
that their use as illuminants for color photography leads to many 
unpleasant surprises, and may be misleading in the selection of 
fabrics, makeup, pigments and the like. 

The need for a direct adequate method of specifying the red, green 
and blue energy content of an illuminant makes it worth while to 
review a proposal put forward a few years ago in connection with the 
spectral sensitivity of photographic materials. Dr. D. R. White, 8 
as a member of the Subcommittee on Sensitivity to Radiant Energy 
of the American Standards Assn., put forward a proposal which is 
extremely pertinent in this connection. Although his proposal was 

4OO 450 

5OO 550 6OO 

Fig. 4. Four monochromatic lines, at 
400, 485, 585 and 700, visually match 
Illuminant C but would scarcely record 
at all on color film at ordinary exposure 

400 450 500 550 60O 65O TOO 

Fig. 5. Transmission curves of stand- 
ard Wratten filters proposed for divi- 
sion of spectrum into appropriate 
zones: dotted line, No. filter; center 
No. 12; right, No. 25. 

limited to monochrome reproduction and to a single illuminant, we 
should like to point out a way in which it could be extended to cover 
color film and color processes, the color balance of light sources, and 
the calibration of correction and compensation filters. 

Since Dr. White's proposal is available in the literature, there is no 
need here for more than a brief summary. Basically, what he pro- 
posed was a simple means of measuring the relative percentages of the 
total sensitivity of an emulsion to Illuminant C, in the red, in the 
green and in the blue, and a simple index number in which three 
values would completely characterize that sensitivity. 

This simple and direct approach involved only the determination 
of filter factors. Thus, if the material required ten times the intensity 
of Illuminant C through a No. 25 red filter to produce a suitable 
standard density that was required with no filter (a filter factor of 10) 


it was evident that the sensitivity in the red was one-tenth of the total 
sensitivity. Similar filter factor determinations in the green and blue 
would assign values for all three zones. However, since the notorious 
inefficiency of green and blue filters would make special correction 
factors necessary, Dr. White adopted the artifice used years before by 
Dr. Eder and used a red, a yellow and a colorless filter, of substantially 
equal efficiency. The unfiltered value minus yellow filtered value 
then gave a total blue reading, and yellow minus red gave the full 
green sensitivity. 

The resulting percentages were expressed in a system of indices 
which it would be pointless to reproduce here, since the steps were 
too great for the differences which are significant in color photography. 
Let us see what happens, however, when we extend the system to light 
sources, color film and corrective filters. 

Dr. White's proposal was to rate the sensitivity of black-and-white 
materials in relation to Illuminant C, which was a logical simplifica- 
tion. Color processes, however, introduce a completely new relation- 
ship between illuminant and sensitized material a reciprocal relation- 
ship between the spectral distribution of energy in the light source and 
the spectral distribution of sensitivity in the film, so that an illuminant 
of the correct color will record balanced densities in the film. 

Since the color of the light source becomes the most important 
variable, in this case, and since it is to the illuminant that corrective 
measures will be applied, it has seemed to us logical to take the illumi- 
nant as the point of departure for the entire system. 

What we have done is, first, to evolve a numerical index which 
accurately describes the spectral distribution of energy in the light 
source, in terms of those attributes which have a bearing on color 
photography. The color film or process is then rated in terms of the 
index of the illuminant which will produce the most nearly neutral 
image of a neutral object. Corrective filters are rated in terms of 
the change which they introduce into the illuminant before it reaches 
the film. Thus, illuminant, filters and film are rated in identical 
units, all derived objectively from easily obtained data. 

The first problem, then, is to find an index which will express the 
relative amounts of energy in three bands of the spectrum, suitably 
measured. For the isolation of these bands, the filters proposed by 
Dr. White seemed eminently suitable, with one modification. He 
proposed to use the Wratten Filter No. 25 for the red, the Wratton 
No. 12 for the yellow, and no filter for the white light exposure, a suit- 
able correction factor being applied to the white light data to simulate 
the filter losses by surface reflection in the yellow and red light expo- 




sures. To eliminate this additional step in computation, it has seemed 
to us more convenient to make the white light measurement through a 
Wratten No. plain gelatin filter. All the data to be presented are 
based on this set of filters, Nos. 25, 12 and 0. It should be pointed 
out in passing that glass filters have been made with very similar 
transmissions, and if it is felt that complete stability of the primaries 
is more important than general availability, a closely matching set of 
glass niters could be produced. Curves of the three gelatin filters 
used are shown in Fig. 5; and in Fig. 6 are shown the effective red, 
green and blue transmissions which result after the subtraction of red 
from yellow and yellow from colorless. As will be seen, the fictitious 
filters which result from this artifice are considerably better than 
actual green and blue filters. The transmission maxima are all high, 

990 600 650 70C 

Fig. 6. Fictitious filter curves which 
result when transmission of No. 12 is 
subtracted from that of No. (left); 
when No. 25 is subtracted from No. 12 
(center); and real curve of No. 25. 

350 400 490 500 

Fig. 7. Curves of Fig. 6 combined 
with spectral sensitivity of a typical 
barrier cell. This gives the proposed 
cutoffs at the long and short-wave 

and the cutoffs are steep. Furthermore, they show a good degree of 
similarity to the curves of leading color processes. 

A special word needs to be said at this point about the boundaries 
of the red, green and blue zones. The boundary between blue and 
green lies at 518, and that between green and red at' 598. This leaves 
the question of the long and short wave limits to be fixed. It is our 
belief that the blue zone should extend to 350, at which point it should 
approach zero, and that the red zone should reach a negligible value 
at 700. These limits are in good agreement with available data on 
color film sensitivities, which we are not at liberty to make public. 
Furthermore, the proposed set of filters, 25, 12 and gives this sort 
of over-all response when used in conjunction with a suitable barrier- 
type cell, as shown in Fig. 7, where filter transmissions are applied 
to the sensitivity of a particular cell used for colorimetric purposes. 


This means that photoelectric measurements may be taken which can 
be expected to show a high degree of similarity to the behavior of color 
film under the same conditions. 

However, all of the computational results presented in this paper 
were obtained with an arbitrary set of primaries which cut off the blue 
at 400 and the red at 675. This was necessary because complete and 
accurate data were not yet available on the band from 350 to 400. 
However, such trials as have been made have shown that this extension 
of the boundaries will not affect the principal results in any material 
way, with the possible exception of one or two arc sources which are 
entirely incidental at this point. 

To summarize, then, the data on which this paper is based are 
derived from an arbitrary set of primaries : 

Blue. A Wratten Filter No. minus a No. 12, with a cut-off at 400. 

Green. A No. 12 minus a No. 25. 

Red. A No. 25, with a cutoff at 675. 

The first step was to compute the energy transmitted by these 
filters for a whole series of black-body radiators, from 2,000 to 12,000 
K, or from 500 to 83 Mireds. Values were computed at 2,000, 3,000, 
3,200, 3,250, 3,400, 4,000, 5,000, 5,900, 6,100, 7,000, 10,000 and 
12,000 degrees. 

For each illuminant, the energy was integrated for all three filters. 
For example, at 164 Mireds (6100 K) the totals were: blue, 1105; 
green, 720; and red, 586. Reducing these to percentages gave: 
blue, 45.8%; green, 29.9%; and red, 24.3%. 

The first intention was to use these percentages as an index, but a 
few trials showed that three digits would have to be used for each 
color to distinguish illuminants with just perceptible color differences, 
so this was abandoned as impractical. 

The system proposed by Dr. White to multiply the percentages 
by 20, then take the logarithm was tried also, but likewise failed to 
distinguish between small steps with a small number of digits in the 

It then occurred to the writers that a ready-made system existed, 
with which a vast number of studio technicians and other engineers 
were already familiar: the decibel system. The decibel is, of course, 
drived by finding the log of a ratio, then multiplying it by 10 if 
referring to a power ratio, or, by 20 if referring to a voltage ratio. 
For the light energy ratios used as example in this article, a ratio 
multiplier of 20 was used. Each color was considered as the ratio 
of its energy to the total energy in the three zones, and this ratio was 
converted to decibels. In the example already given, the percentages 


worked out as follows: 45.8%, 13.22 db; 29.9%, 9.51 db; and 24.3%, 
7.71 db. 

Since we are interested only in ratios, and not in absolute energy 
levels, and since the three always add up to 100% or unity, it was 
obvious that two of the values would be enough to describe the light 
source. The red value was, therefore, subtracted from all three 
decibel values, reducing red to zero and the others to proper relative 
levels. Numerous trials had shown that the second figure after the 
decimal point could be dropped without loss of the specified accuracy, 
so the adjusted values became: blue, 5.5; and green, 1.8; with 
"red 0.0" implied. This was expressed as an index in the form 5.5/1.8, 
which immediately tells us that in comparison with red, blue is 
"up" 5.5 db and green is "up" 1.8 db. In other cases, of course, 
either or both may be "down" as compared to red, in which case 
the db values are preceded by a minus sign. For want of a better 
name, we call this Spectral Distribution Index, or SDI. 

-18 -16 .14-12-10-8-6-4-2 O 2 4 6 8 10 12 

Fig. 8. Straight-line locus which re- 
sults when decibel values of blue-red 
ratio (horizontal axis) and green-red 
ratio (vertical axis) are plotted against 
each other, 2,000 to 12,000 degrees 
Kelvin. Mired values (below the locus ) 
are spaced linearly for the limits within 
which Wien's law is valid. 

However, as an example of the results obtained with one set of 
primaries, the graph is shown in its entirety in Fig. 8 and with certain 
areas on a larger scale in Figs. 9 and 10. 

As regards the Spectral Sensitivity Index, or SSI, to be applied to 
the film, it seems to us that the simplest procedure is to describe the 
film in terms of light to which it is balanced. Thus, a color film -bal- 
anced to an illuminant with an SDI of -4.9/-2.0 would have the 
same figures as its SSI. Actually, of course, the sensitivities are 
reciprocal, but we are not interested in sensitivities per se, but only in 
their equilibrium with a certain spectral energy distribution. 

Calibration of correction filters is greatly simplified by the use of 
the decibel concept. A filter is rated in terms of the decibel reduction 
which it effects in each zone, and this figure is arrived at by multiply- 
ing the filter density in the blue, the green and the red by 20. Since 
this density may be measured on a color densitometer, or computed 
from an accurate spectral curve, calibration becomes an extremely 
simple matter. 




We would suggest, however, that the Spectral Absorption Index, 
or SAI of the filter, should retain all three values rather than be 
reduced to a form which eliminates the red. In this form, the gray 
content, or density common to all three colors, will be evident, and 
the index will show both the change in color of the light and the ex- 
posure increase, if any. 

If, for example, a filter with an SAI of 1.0/0.5/0.5 is used with an 
illuminant of index 6.0/1.8, we should make the following simple 
computation : 

6.0 1.8 0.0 

1.0 0.5 0.5 





Fig. 9. Enlarged section of the db 
locus, taking in commonly used in- 
candescent lamp Kelvin temperatures; 
as might be expected, the point repre- 
senting the light from a low-intensity 
carbon arc is off the black-body locus. 

Fig. 10. Enlarged section of db locus, 
taking in sources of higher temperature : 
A, studio broadside; B, 170 M-R with 
Y-l filter; C, average noon sunlight; 
D, Technicolor unit with Whitelite 
6300 filter; E, daylight falling on 
horizontal plane, fairly clear; F, same 
on a clear day; G, sun outside earth's 
atmosphere; H, Graf A. C. high- 
intensity arc; I, complete overcast; J, 
Technicolor unit with Whitelite 7100 
filter; K, Illuminant C; L, sunshine 
white flame arc; M, north sky light on 
a clear day. 

Fig. 11. Spectra Color Temperature 
Meter fitted with the proposed new 
decibel scales indicating the blue-red 
and green-red ratios. 


Adding 0.5 to all three, to bring red back to its zero level, we have: 

5.0 1.3 -0.5 

0.5 0.5 0.5 

5.5 1.8 0.0 

This reckoning, which would usually be carried out mentally, tells us 
that the illuminant has been reduced to an SDI of 5.5/1.8, or, in other 
words, Illuminant C has been reduced to 6100 K. The total energy 
loss is 1.0 + 0.5 + 0.5, or 2 db, which must be made up by a suitable 
exposure increase, 6 db corresponding to one full stop. 

For purposes of quick comparison, and to illustrate the concise 
manner in which information is conveyed, it may be mentioned that 
while 6100 K is 5.5/1.8, Illuminant C is 6.0/1.8, which tells us at once 
that, in the latter, blue is up another half a db and that green is identi- 

Having derived the index values for the selected series of black-body 
radiators, the next step was to plot them on a graph. This was 
carried out, with green db values along one axis and blue values along 
the other. 

When this had been completed, a very interesting and useful prop- 
erty of the index emerged : the locus connecting all of the points was 
a straight line, with a completely linear scale of Mired values. This 
meant, of course, that least perceptible color differences were a sub- 
stantially constant linear amount from 500 to 91 Mireds, or 2,000 to 
11,000 K. The linear scale and the perfect straightness of the locus 
make it both easy and safe to interpolate values, and probably to 
extrapolate as well. Two computations serve to establish the locus, 
and a third to verify it. 

Nonblack-body radiators will, in general, fall at points off the line. 
Those which fall on the line may be considered, for photographic 
purposes, as black-body radiators. 

There would be little point, at this time, in publishing the decibel 
values for all of the illuminants studied, since this work will be re- 
peated in the near future with a corrected set of primaries. This may 
affect the gradient of the locus, and the reference level, but there is no 
reason to anticipate any change in the over-all relationships which 
have been established. 

All of the foregoing can be applied by the manufacturer, who could 
mark the film and filters with appropriate values, but even without 
this there would be little difficulty in the application of the system 
by a single user. The foregoing procedures can easily be correlated 
with standard color densitometric procedures. The laboratory could 


advise the cameraman as to the best illuminant for a particular emul- 
sion, and if the cameraman, under the pressure of production were 
obliged to shoot scenes with an incorrect illuminant, a single dot on a 
graph would tell the laboratory the nature and amount of the devia- 
tion to be expected. 

An instrument incorporating the logarithmic scales described here 
for the blue-red and green-red values is shown in Fig. 11. 

Much work remains to be and will be done on the system herein 
described. In the meantime, it has seemed to us that the results are 
sufficiently promising and interesting to warrant publication and 
availability for discussion at this time. 


1. ASA Z38.21 ( 1947), "American standard method for determining photographic 
speed and exposure index," American Standards Association, 70 E. 45 St., 
New York. 

2. E. J. Wall, Practical Color Photography, Amer. Phot. Publishing Co., p. 59, 
Boston, 1928. 

3. "A Maxwell Triangle Yielding Uniform Chromaticity Scales," Research 
Paper RP756, Bureau of Standards, p. 52. 

4. Norman Macbeth and Dorothy Nickerson, "Spectral characteristics of light 
sources," Jour. SMPE, vol. 52, pp. 157-183, Feb. 1949. 

5. R. M. Evans, An Introduction to Color, pp. 213-14, John Wiley, New York, 


6. Committee on Colorimetry, "Physical Concepts; radiant energy and its 
measurement," /. Opt. Soc. Amer., vol. 34, pp. 183-218, Apr. 1944. 

7. L. A. Jones and H. R. Condit, "Sunlight and skylight as determinants of 
photographic exposure, Part II, Scene structure, directional index, photo- 
graphic efficiency of daylight, safety factors and evaluation of camera ex- 
posure," /. Opt. Soc. Amer., vol. 39, pp. 94-135 (p. 123), Feb. 1949. 

8. "A Spectral Sensitivity Index for photographic emulsions and calculations 
based thereon," Jour. PSA, vol. 9, No. 8, p. 386, October, 1943; and No. 9, 
p. 585, December, 1943. 

The Stroboscope as a Light Source 
For Motion Pictures 




SUMMARY: The stroboscope has long been proposed as a source of illumina- 
tion for ordinary motion pictures because of several attractive technical fea- 
tures; however, as has been pointed out clearly by F. E. Carlson 1 recently, 
there are serious disadvantages that must be overcome before the flashtube 
finds widespread practical use for everyday motion picture studio photogra- 
phy. This study reports additional experiences with flashtubes especially at 
large power input as would be needed in picture taking. One object was to 
find the upper power limitations of existing commercial flashtubes. A fur- 
ther object was to study the design and performance of a three-phase ef- 
ficient power supply. 

MOTION PICTURE studio lighting has been a challenging problem 
for electrical engineers especially since the advent of sound 
and color photography. There is still need for improved light sources 
and it is this urge which has prompted the effort reported in this 

The theory of flashtubes and circuits has been described in several 
articles given in the bibliography 2 " 4 and therefore will not be re- 
peated here. 

First, it is in order to discuss briefly the advantages that are in- 
herent in the stroboscopic system of illumination : 

1. The flashes of light occur only while the camera shutter is open, 
resulting in 100% utilization of the light. A conventional camera 
with continuous light uses only about 50% of the light. Thus a 
doubling of efficiency is possible. 

2. The light-producing efficiency of Xenon-filled electronic flash- 
tubes is higher than tungsten lamps. 

3. The effective color temperature of Xenon tubes is almost the 
same as for daylight. Therefore the same camera and film equipment 
can be used for outdoor and studio photography. 

The main disadvantages of the stroboscopic system are : 

PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. This is 
a condensation from a Master's thesis at the Massachusetts Institute of Tech- 
nology, Electrical Engineering Dept., by R. S. Carlson. 


1. The flicker of the light at 24 cycles (one flash per frame) is un- 
bearable by the actors. It has been proposed that multiple flashes 
be used to avoid this difficulty. Another possible solution is to use a 
combination of continuous and flashing lights. 

2. The short flash of the stroboscopic source produces a clear sharp 
photograph of a rapidly moving subject instead of the desired blurred 

3. There is a certain amount of noise associated with a powerful 
stroboscopic source that may be objectionable on sound stages. 
Acoustical treatment will be necessary for the tubes. 

This paper is concerned with the design of powerful 24-cycle strobo- 
scopic sources using existing tubes with the thought that the advan- 
tages of the system will make the system useful, regardless of the dis- 
advantages, for special applications. 


The .practical upper power limit for a flashtube operated as a strobo- 
scope is considerably different than when operated for single-flash 
photography. Some of the important factors influencing tube de- 
sign and use are discussed in the following. 

Flashtubes when used for single flashes at remote intervals of time 
are not concerned with the over-all temperature of the tube. The 
inner surface of the tube is influenced by the transient temperature 
pulse but the average temperature does not rise to a point where 
performance is influenced. One of the upper limits of energy that 
can be put into a flashtube in a single flash depends upon the surface 
temperature conditions. With a glass tube, an overload will result 
in a crazed surface consisting of a network of surface cracks. With a 
quartz tube, the overload will be several times greater for the same 
internal dimensions and will produce a cloudy appearance which is 
apparently caused by condensed quartz vapor that has been evapo- 
rated by the energy from the flash. Experiments show that the light 
output is not materially reduced for both glass and quartz tubes even 
after serious crazing or cloudiness; however, the life of such tubes 
may be greatly reduced. 

It is very important in single-flash work to load the flashtube as 
high as possible in order to enjoy the resulting high efficiency; there- 
fore an effort is made to load single-flash tubes to a maximum. 

For continuous stroboscopic use, it is not possible to load the 
tubes since the average temperature of the tube walls will become ex- 
cessive. A glass tube exhibits wall conduction when it becomes hot. 
The trigger electrode potential in some cases will cause a puncture of 




the hot glass wall allowing the entrance of air and thereby ruining the 
tube. In other cases the conduction by the hot glass effectively short 
circuits the spark excitation, resulting in flash missing. The quartz 
tubes exhibit similar characteristics except the failures exhibit them- 
selves at a much higher temperature. Puncture of a quartz tube by 
the excitation is a rather rare event compared to skipping, while with 
a glass tube the opposite is the general case. A skipping quartz tube 


Fig. 1. General Electric quartz flash- 
tube No. FT-417. 

Fig. 2. General Electric quartz 
flashtube No. FT-623. 

is usually not damaged. It can be made to operate satisfactorily by 
reducing the power input or by artificial cooling with air or water. 

Another limiting factor for stroboscopic tubes is electrode tem- 
perature. For single-flash applications the electrodes can be small; 
however, for stroboscopic use, the area must be increased to radiate 
the continuous electrode losses. 

Flashtubes such as the FT-503 and FT-524 (General Electric Co.) 
have a quartz spiral but small electrodes limit the input to about 


700 w since with that input the electrodes reach a yellow heat. Addi- 
tional continuous inpiit will result in excessive electrode evaporation 
and other objectionable characteristics. A flow of air is required 
also for 700-w input to cool the quartz spiral and the FT-524 tube is 
especially designed for this with an open end so that a draft of air can 
be forced directly through the tube. These tubes were not studied 
since their output was so small. 

Two quartz flashtubes are available which have a larger electrode 
assembly which is capable of handling greater power. These tubes 
are the FT-617 and the FT-417 (Fig. 1), both General Electric Co. 
products. Flashtube No. FT-617 is the same as No. FT-623 (Fig. 
2) except that the glass envelope is open to facilitate cooling. An 
uncoiled version of the FT-417 is also available from General Elec- 
tric Co. and is identified as the FT-427. Approximate dimensions 
of the helical tubes follow: 

Helix Helix Number Tubing, outside 

diameter, in. length, in. of turns diameter Gas 

FT-417 iy s 1% 4 M Xenon 

FT-617 2^6 35^ Xenon 

The electrodes of both tubes consist of a sintered pellet of tungsten, 
nickel and barium, welded to a J^-in. solid iron post abeut 1 in. in 

The actual tubes used in our tests were an earlier variation of the 
FT-617 and had a spiral of six turns instead of five. Likewise the 
FT-417 tube used here was an early experimental type which may 
be slightly different from production tubes. However it is thought 
that the results from these tubes will be comparable to the tubes 
currently available. No experiments have been made to determine 
the life under the conditions reported here. Our results should be 

Table I\ Efficiency Data 







lumen- sec 


FT-214 Std. 








considered to be preliminary and should be checked with production 

The efficiency data for the two tubes under conditions as used in 
stroboscopic circuits are shown in Table I. 

It has been found that each given type of tube can be operated at 
some experimentally determined maximum power input for given 
cooling conditions. In terms of the circuit this power is approxi- 


P = - - / watts input 

where C = capacitance in farads, 

E = voltage to which the capacitance is charged prior to 

the flash in volts, and 
/ = the frequency of flashing in flashes per second. 

Once the maximum power is known and the frequency selected, 
then the watt-seconds loading CE 2 /2 is fixed. Likewise the efficiency 
is determined as given by the characteristics of the flashtube. 

An important conclusion from the above that bothers the designer 
of a stroboscope is the decrease of possible efficiency with an increase 
of frequency when the power is constant. The importance of the 
use of quartz and forced cooling is apparent since both permit a 
higher power input and therefore the tube operates at a higher ef- 

Without forced cooling, both the FT-617 and the FT-417 are limited 
in continuous stroboscopic operation by the heating of the quartz 
tubing. The FT-617 with its large-area coil structure is easy to cool 
with a blast of air through the tube. The FT-417 does not cool so 
easily since the air flow is not uniform around its small coil. Hot 
spots tend to develop on the opposite side of the tube from where the 
air is blown. These can be seen since the quartz reaches a red heat. 

The FT-617 tube cannot be operated with a power input greater 
than 4 kw even with an unlimited air blast since the electrodes reach a 
temperature where holdover tendencies are in evidence. Further- 
more the electrodes tend to evaporate and discolor the tube. 

The FT-417 is likewise limited to 4 kw because of the electrode 
structure. Cooling of the quartz spiral on the FT-417 is more of a 
problem than on the FT-617 because of difficulty of getting adequate 
heat transfer. A few experiments were made with the FT-417 coil 
immersed in water. The operation was satisfactory even with 10-kw 
input for short bursts with electrode heating again being the limiting 




Intermittent use of a stroboscope can be accomplished at higher 
power inputs than when used continuously. For example an earlier 
version of the FT-617 tube discussed later which operated contin- 
uously at 4-kw input could be run for 30 sec with 10 kw before the 
electrodes reached an excessive temperature. 

The design of a power supply to operate a stroboscope tube with 
an input of 4 to 10 kw requires some special considerations. For 
example, if a single-phase a-c circuit is used, the power supply would 
need to have filter capacitors to smooth out the ripples which might 
cause nonuniformity of the light flashes. 

The preferred method of supplying power from an a-c distribution 
system for power values over about a kilowatt is the use of the con- 
ventional three-phase system. This type of system was used for a 

3 </> IHPUT 

Fig. 3. Three-phase power supply to operate a flashtube 
at 10-kw input. 

power source here with a six-phase rectifying system which has a 
rather small ripple. In fact with the choke charging system shown, 
there is no need for a capacitor filter. 

Figure 3 shows the diagram of the power supply which was used. 
This 1000-v supply can deliver 10 kw to a stroboscope lamp at 2000 v. 

Power can be purchased from the power company at the proper 
voltage (about 850 v) to supply the rectifiers and flashtube. In this 
way the cost of transformers is eliminated from the power supply 
equipment at a considerable saving in expense and weight. 

The six-phase rectifier system reported here shows mercury vapor 
rectifiers, General Electric Co. type FG-32. Serious consideration 
should be given to selenium rectifiers for this service since the fila- 
ment-heating delay complications would be eliminated. At present 
the selenium system would cost more than tubes but would have 
operational advantages that might justify the extra investment. 

The size and thereby the cost of the charging inductor is deter- 


- A 

Fig. 4. Voltages and cur- 
rent in charging circuit dur- 
ing flashtube operation at 24 
flashes/sec. L = 1.62 h; 
C = 46 /if- 

KEY to Figs. 4, 5 and 6: 

Time scale: 1 in. = 0.08 sec. (See legend beside Fig. 4 above.) 

A. V c (t), storage capacitor and flashtube voltage, 1 in. = 3,780 v. 

B. E 8 (t\ rectifier output voltage, 1 in. = 3,000 v. 

C. Vi(t\ charging inductor voltage, 1 in. = 3,220 v. 

D. i(t), charging current, 1 in. = 19 amp. 

mined by the transient watt-second storage capacity. As will be 
shown later this energy is one-fourth that of the discharge capacitor. 
In our preliminary design investigations, we found that the inductor 
might weigh more than the capacitor. A further engineering study is 
required to arrive at the best design of the inductance. 

The power transformer, if used, can be built with leakage reactance 
that can supply all or part of the inductive component of the circuit. 

The flashtube energy storage capacitor C (Fig. 3) is charged to ap- 
proximately twice the rectifier output voltage, through the charging 
inductor L. The capacitor holds the voltage because the rectifiers 
prevent a current reversal until the flashtube is triggered. When the 
flashtube is ionized, it breaks down and discharges the capacitor to 
the relatively low voltage at which the flashtube becomes non- 
conducting. The capacitor then recharges through the inductor, 
completing the cycle. 

Figure 4 shows oscillographic records of the current and voltage 
relations in the charging circuit during repetitive operation at 24 
flashes/sec, with a storage capacitor of 46 ;uf (microfarads) and a 
charging inductor of 1.62 h (henries). The rectifier output voltage 
has a ripple with six times the frequency of the supply voltage, caused 





Fig. 5. Voltages and current in charging circuit, showing flashtube hold- 
over during first pulse of charging current. L = 1.62h; C = 94juf. 


Fig. 6. Voltages and current in charging circuit, showing flashtube hold- 
over during second pulse of charging current. L = 1.62 h; C = 94 /if. 

by the full-wave, three-phase rectification. The first pulses of cur- 
rent and voltage are larger than succeeding pulses, because the first 
charging cycle starts with the capacitor initially uncharged. The 
slight drop in the capacitor voltage after it reaches a maximum is 
caused by the drain of the oscillograph element and multiplier. 


A mathematical analysis of the charging circuit yields the expres- 
sions given below, in which the resistance of the charging circuit has 
been considered negligible. For a general analysis of d-c charging, 
see Pulse Generators. 5 

sn V ' 


V e (t) = E.+ [V c (0) - E s ]cos Vl/LC t (2) 


i(t) is the charging current as a function time, t, 

V e (t) is the storage capacitor voltage, 

E 8 is the rectifier output voltage during charging, 

F c (0) is the capacitor voltage at the beginning of a charging cycle, 

L is the charging inductance in henries, 

C is the energy-storage capacitance in farads, and 

t = is the beginning of a charging pulse. 

Equations (1) and (2) are valid only between ^ t ^ 7r\/ZC; that is, 
the equations are valid for the duration of the charging-current pulse, 
or while the charging current is positive. When the charging current 
is zero, the drop across the inductor is zero and the rectifier voltage 
becomes equal to the capacitor voltage, which is about twice E s , and 
Eq. (1) and (2) no longer hold. If the current has not become zero 
by the time the flashtube is fired, it is likely that continuous conduc- 
tion of the flashtube will result, with current being supplied directly 
from the power supply to the flashtube. To assure that the current 
will be zero at the time of the flashtube firing, *\/LC should be less 
than the period between successive flashes. However, making the 
charging period very short means higher instantaneous charging 
current, which makes the duty harder on the rectifier tubes and other 
parts of the power supply. The waveform of the input alternating 
current also becomes poorer as the charging time becomes a smaller 
fraction of the period between flashes. 

From Eq. (1) and (2), it is noted that the maximum value of charg- 
ing current is 

_ E. - VM (3} 

and the maximum capacitor voltage is 

2J!.-y.(0) (4) 


Equations (3) and (4) are useful in predicting two quantities of pri- 
mary concern in the design of the power supply : maximum charging 
current and maximum or final capacitor voltage. 

From Eq. (3) it is obvious that the maximum current should vary 
inversely as the square root of the charging inductance. Equation 
(4) indicates that the maximum capacitor voltage is not affected by 
changes in charging inductance. A change in the storage capacitance 
affects the maximum current the inverse of the way a change in 
charging inductance does; that is, the maximum charging current in- 
creases directly as the square root of the storage capacitance. 

In the charging circuit, the inductor must be capable of storing the 
energy LP maK /2. Changing the size of the inductance will not alter 
the required energy-storage capacity of the inductor, since 7 max in- 
creases inversely as the square root of L. The required energy-stor- 
age capacity of the inductor is 

[E.- V.(0) Hi CV\^ 

'- [ZET^VM] ~ 

From Eq. (5) it is seen that on the initial charging pulse, since V c (0) 
is zero, the peak energy storage in the inductor is one-fourth of the 
final energy storage in the capacitor, but is less than one-fourth on 
succeeding pulses, when V e (0) is not zero. 

To summarize the requirements of a charging inductor: 

1. The peak energy-storage capacity of the inductor must be at 
least one-fourth the final energy storage of the capacitor. 

2. For a given amount of capacitance, the inductance should not 
be so large that ir\/LC exceeds the period between flashes. 

3. The inductance should not be so small that the charging time is 
too short in comparison with the period between flashes. Low induct- 
ance, besides increasing the severity of the duty on circuit elements, 
is less effective in isolating the flashtube from the power supply im- 
mediately following the flash. 

Besides the above considerations, an economical inductor design 
must take into account the proper balance of copper, iron, air gap and 
insulation. Of course, if one inductor is to be used over a range of 
flashing rates or capacitances, optimum design over the whole range 
is not possible, and some compromises must be made. 

The oscillograms of Figs. 5 and 6 illustrate two conditions that 
lead to holdover, or failure of the flashtube to stop conducting. In 
Fig. 5, the flashtube was flashed before maximum voltage and zero 
charging current occurred. Since there was a current flowing in the 
inductor at the time of flashing, it continued to flow through the flash- 


tube, being limited only by the charging inductance, until finally the 
circuit breakers opened. 

In Fig. 6, the current had gone to zero, following the charging of 
the capacitor to full voltage. The flashtube was flashed, and the 
normal build-up of current and capacitor voltage had begun when 
apparently the flashtube "broke down" again, and a holdover similar 
to the one in Fig. 5 occurred. The breakdown of the flashtube was 
probably caused by failure of the inductor to provide isolation from 
the power supply long enough for complete deionization of the flash- 
tube to take place. With the FT-617A, it was impossible to prevent 
holdover by increasing the size of the inductance when the input was 
above about 4 kw at 24 flashes/sec, with flashtube voltage at 2,000. 
The inductance was increased to the limit at which ir^/LC was just 
below the period between flashes, and holdover still occurred occasion- 

With the FT-417, no trouble was experienced with holdover even 
when loaded to 10 kw. This is attributed largely to the fact that the 
smaller tube had a much shorter deionization time than the FT-617A. 
Deionization after conduction is the result of ion and electron recom- 
bination mainly on the inner surface. Since the ratio of surface area 
to volume increases when diameter decreases, the smaller tube should 
become deionized more readily. 

Two related problems that are of considerable importance are 
those of flashtube cooling and flashtube starting, or triggering. When 
the flashtube becomes too hot, the insulating property of quartz be- 
comes very poor, and the starting pulse instead of ionizing the gas 
is shunted around the gas by the quartz tubing, and the flashtube fails 
to fire. With the FT-617, this is easily remedied by using a small air 
blower for forced air-cooling. Whereas the flashtube could be oper- 
ated without cooling at a power input of 1 J/ kw for 1 min, or at 4 kw 
for 30 sec, with forced air-cooling it could be operated continuously 
with an input of 5 kw, except for the holdover problem at this input. 

The smaller FT-417 heated up more rapidly and would not operate 
for more than about 20 sec at 5-kw input, even with very strong 
forced air-cooling. 

An interesting experiment in water-cooling was performed with 
the FT-417. It ran smoothly with an input of 10 kw while the whole 
flashtube, except for the main electrode lead-in wires, was submerged 
in water. The flashtube would apparently fire regularly for as long as 
desired. However, the electrodes became white-hot in about 30 sec 
at this input. After a total flashing time of about 2 min at this in- 
put, the flashtube showed several signs of hard use. The electrodes 


were blackened and pitted, and some of the metal had sputtered to 
the adjacent quartz tubing. The inside of the quartz tubing was 
clouded considerably, presumably caused by a melting or vaporiza- 
tion of the quartz on the inside of the tubing. This experiment 
indicates the possible practicability of using a water-cooled flash tube, 
though of course there would be numerous problems connected with 
the design of such a flashtube. 


A major problem with the use of stroboscopic sources when people 
are illuminated is the flicker effect. Intermittent 24-cycle light is 
very disturbing. 

A brief investigation was made to determine the possibility of reliev- 
ing the flicker effect by supplementing the stroboscopic light with 
some continuous light. It was found that the ratio of continuous 
light from a tungsten lamp to stroboscopic light, for almost complete 
elimination of the flicker effect, was about 10 to 1. For a basis of 
comparison, the light output in lumens for the stroboscopic source 
was taken as 24 times the lumen-seconds per flash. With this ratio of 
continuous to stroboscopic light of 10 to 1, the usually observed 
stroboscopic effects such as the apparent change in speed or direc- 
tion of rotation of wheels, jerkiness of movements, and flickering of 
light were hardly detectable. However, at ratios of about 8 to 1 
and lower, the stroboscopic and flicker effects were only slightly 
less than when only the stroboscopic source was used. 

The problems of flashtube noise and the excess blue of the flash- 
tube light output will also have to be met in using flashtubes in movie 
work. Some noise reduction may be obtained by tube design and by 
using an inductance in series with the flashtube. For color correc- 
tion for daylight-type Kodachrome film, a CC13 or CC15 filter is 
recommended. It may be that a proper mixture of stroboscopic 
lighting and continuous lighting would make the use of filters un- 

Another method of reducing flicker is to increase the flashing rate. 
To do this reduces the efficiency as has been pointed out before. 


The upper limit of power input to a flashtube is determined by the 
thermal capacity of the flashtube. If the quartz tubing becomes too 
hot, the flashtube fails to fire and the tubing may be damaged; if the 
electrodes become too hot, they may be damaged and the flashtube 
tends to hold over more readily. Which of the factors is the limiting 


one depends upon the tube and electrode design, the circuit design, 
the method of cooling, and the length of time that operation is de- 

Once the maximum power input is determined for a given tube, 
circuit and cooling, it will be essentially constant regardless of fre- 
quency. Then, since power input is a function of input per flash and 
flashes per second, the permissible input to the flashtube for each 
flash will depend upon maximum power input and the rate of flash- 
ing, or CF 2 /2 = P max /y watt-seconds per flash. 

The stroboscopic flashtube unit described here when operated with 
the FT-617, with a storage capacitance of 74 juf charged to 2,000 v, 
will operate continuously at a flashing rate of 24 flashes/sec, with 
forced air-cooling from a 35-w, 8,400-rpm blower; operation for 
about 30 sec is possible with no forced air-cooling. The power input 
to the system is about 3.6 kw, and the energy per flash is about 150 

The light output is 2,600 Im-sec/flash, since the efficiency is about 
17 Im/w. If the flashtube is placed in a reflector with an efficiency 
of 10, it produces 104 Im/sec/sq ft at a distance of 5 ft. This amount 
of light is sufficient for photography with daylight-type Kodachrome 
film at //3.5, and considerable coverage could be obtained with 
several such flashtubes. However, it is obvious that increasing the 
flashing rate means cutting the watt-seconds per flash in order to 
stay within the power rating of the flashtube. For example, if the 
flashing rate in the above case were increased to 48 flashes/sec, the 
energy input per flash would have to be cut in half, or to 75 w-sec/ 
flash. The light output would be cut by more than half, since the ef- 
ficiency would be lowered. 

Thus it seems that in order to get a flashtube source more useful for 
movie photography, it will be necessary to increase the thermal ca- 
pacity of flashtubes so that they can be operated repetitively at 
higher loadings and higher efficiencies. One possibility is a water- 
cooled flashtube. Another is a tube with large electrodes. 


1. F. E. Carlson, "Flashtubes: a potential illuminant for motion picture photog- 

raphy?" Jour. SMPE, vol. 48, pp. 395^06; May, 1947. 

2. H. E. Edgerton, "Photographic use of electrical discharge flashtubes," /. 

Opt. Soc. Amer., vol. 36, p. 390; July, 1946. 

3. F. E. Carlson and D. A. Pritchard, "The characteristics and applications of 

flashtubes," Ilium. Eng., vol. 42, p. 235; February, 1947. 

4. H. M. Lester, "Electronic flashtube illumination for specialized motion pic- 

ture photography," Jour. SMPE, vol. 50, pp. 208-232; March, 1948. 

5. G. N. Glasoe and Jean V. Lebacqz, MIT Radiation Laboratory Series, No. 

5, Pulse Generators, 728 pp., McGraw-Hill, New York, 1948. 

Study of Sealed Beam Lamps 
For Motion Picture Set Lighting 



SUMMARY: Limitations and advantages of sealed beam lamps for motion 
picture set lighting are disclosed. Comparisons of light output and distribu- 
tion vs. regular studio spot lamps are made; also, lamp efficiency, life and 
weight of equipment are discussed. Lamp requirements for motion picture 
set lighting are presented and the methods followed by the Motion Picture 
Research Council to determine possible usage of sealed beam lamps are de- 

FOR SOME TIME, many people associated with motion picture 
set lighting have thought the incandescent lighting equipment 
now being used appeared heavier and more bulky than necessary. 
The recent strict budgets have accelerated the exploration for a simple 
light source and flexible lighting equipment. Attention was drawn 
to the sealed beam type of lamp having a built-in reflector. This type 
of lamp is presently being used primarily for fill light on locations and 
was used for television studio lighting as early as July, 1939. l 

The amount of illumination obtained from this type of lamp, such 
as the reflector photoflood (RFL-2) and reflector spot (RSP-2) (see 
Fig. 1), is very impressive. Such lamps have .an average life of 6 
hr and dissipate 500 w. This lamp construction and operation will 
supply more light per watt to the studio set than present spot lamps, 
but the fixed focus design sacrifices the ability to spot or flood the light 

The Colortran Converter Co., Hollywood, Calif., supplies to the 
industry portable kits in which long-life industrial lamps are used. 
These lamps are operated at an elevated voltage by transformer action. 
So operated, these industrial lamps have a light output, color tempera- 
ture and life comparable to photofloods. Such tactics increase lamp 
efficiency since light output increases approximately twice as fast as 
the wattage when the melting point of the tungsten filament is 
approached (see Fig. 2). 

This Colortran equipment is very popular with the studios, espe- 
cially for location work where close shooting quarters, power consump- 

1 William C. Eddy, "Television lighting," Jour. SMPE, vol. 33, pp. 41-53, July 

PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. 







Fig. 1. Reflector Spot (RSP-2) and Reflector Flood (RFL-2); 500-w, 6-hr life. 













^ 20 









> r 





, -* 







100 120 140 160 


100 120 140 160 






u 3100 


" 2900 


100 120 140 160 


Fig. 2. Characteristic 
curves of standard 120-v, 
1,000-hr life, incandescent 
lamps when operated at other 
than the rated voltage. 


tion and portability are important considerations. The industrial 
150-w PAR-38 flood (Fig. 3) appears to be the most suitable lamp for 
this equipment. 

Due to the interest shown in reflector-type lamps, the Motion 
Picture Research Council studied the characteristics of such lamps 
to see if over-all advantages existed which would make them more 
economical for studio use. Existing built-in reflector lamps were 
measured and compared with present equipment (see Table I). 
Light measurements were made at the center, edge and 6 deg outside 
a circle 8^ ft in diameter, located 20 ft from the light source. 

It is interesting to compare the 500-w reflector flood (RFL-2) 
operating at 120 v with the PAR-38 flood operating at 185 v. Under 
these operating conditions, the lamps have comparable life. Inside 
the circle, the PAR-38 delivers twice as much light as the RFL-2. 
The RFL-2 has a more uniform field of illumination, but the fall-off 
of the PAR-38 is not considered too serious. Outside the circle, the 
RFL-2 delivers considerably more light. In general, however, this 
outside illumination is undesirable and will be masked off with 
barndoors* or gobosf unless the lamp is being used to supply fill light. 

Photographic tests were made at Paramount Studios to determine 
whether acceptable black-and-white picture quality could be obtained 
using only reflector-type lamps. Colortran transformers were used 
for over-voltage operation. Identical long shots and close-ups with 
the same light key and at the same lens stop were taken; first with 
standard lighting equipment and then repeated using sealed beam 
reflector-type lamps only (Fig. 4). Well-known cinematographers, 
professional players, a dressed set and the recording of sound were 
used. Equivalent and highly satisfactory photography was obtained 
using barndoors, scrims and gobos, in spite of the fact that sealed 
beam lamps are diffused light sources. The sealed beam lamps re- 
quired approximately one-third the wattage of that of standard 
lighting. Less time was required to make setups using the sealed 
beam lamps. However, a fair comparison cannot be made because 
the studio lamps were set up first and lamp placement had already 
been determined when the sealed beam lamps were set up. Individual 
dimming was supplied to each sealed beam lamp through a small vari- 
able resistor. The sealed beam lamps had a higher color temperature, 
although the foot-candle reading was kept the same for both types 

* Black, movable extension doors hinged on lamp housing to restrict light from 

reaching certain areas, 
t Normally, cloth masks of various sizes, mounted separately from lamp housings, 

used to block light from certain areas. 



TABLE I. Test Results of Light Output 

Light Distribution* 


Color Center Edge 2 Ftf 
Life Temp., foot- foot- outside 
Volts Watts Hours Kelvin candle candle circle 

From Commercial Sealed Beam Lamps 
































































1 A 



































































































1000 R-80 









1000 R-80 









1000 R-80 









Present Studio 


for comparison 

500 MP 








500 MP 








500 CP 








750 MP 








750 MP 








750 CP 








1000 MP 








1000 MP 








2000 MP 








2000 MP 








2000 CP 








See notes on following page. 


of lighting. This increase of energy in the blue region of the spectrum 
increased the actinic value so that a negative of higher density was ob- 
tained with the sealed beam lamps. Thus, to match print density 
from the two negatives, it was necessary to print two light steps higher 
with the negative using sealed beam lamps. Rigging and striking 
time was reduced because the lamp housings weighed only a fraction 
of the weight of standard equipment. 

With evidence that satisfactory photography could be obtained 
with sealed beam lamps, the lamp manufacturers were requested to 
provide special lamps for experimental work. Special filaments in 
existing bulbs of various sizes were supplied. These lamps were life 
tested and the rate of aging was measured for different filament orien- 

Of these experimental lamps, the PAR-38 flood bulb with a 115-v, 
500-w, CC-6 filament seemed the most desirable for our application. 
Life tests indicated that an average life of 6 hr could be expected. 
Damage of the reflector was observed directly about the filament dur- 
ing life tests, but had no serious effect. The lamp delivered over 
three times the amount of light of a 500-w RFL-2 in the 8^-ft-diam- 
eter circle and approximately 75% of the light of a 2,000-w studio 
Junior spot lamp. 

To realize fully the possible advantages of sealed beam lamps for 
studio use, the Research Council designed housings, associated 
equipment and scaffolds to accommodate PAR-38 or R-40 bulbs. 
Figure 5 shows the lamp housing and the adjustable scissor bracket 
in the retracted position. Figure 6 is a multiple exposure showing 
possible positions of the scissors. Figure 7 is a small set rigged with 
this special lamp equipment. Figure 8 shows the type of scaffold 
used, which is supported entirely by the set walls. 

Individual electrical dimming of the lamps can be done either from 
a master control board located on the stage floor, or with a rheostat 
located on each housing and operated at the lamp. Remote control 
on a master board affords quick lamp adjustments for the gaffer (chief 
set electrician), but entails the numbering of each lamp and the run- 
ning of cables from the lamps to correspondingly numbered controls 

NOTE: Underlined figures of life and color temperature are calculated for an average 

lamp and subject to considerable variation for an individual lamp. 

* Measurements made in standard studio lamp housings flooded to 20% reduction 

in lamp beam candlepower at edge of circle 8^ ft in diameter, 20 ft from lamp; 

this is to simplify comparison with reflector lamps, 
t Airplane landing lamp, 
j 6 from center. 

Fig. 3. PAR-38 Flood; 150-w, 1,000-hr life. 

Fig. 4. Master long shot using sealed beam lamps. 




at the master board. Having a rheostat on each lamp housing pro- 
vides a much simpler wiring setup, but requires manual dimming at 
the lamp by the operator. Choice between these two methods will be 
determined largely by operating experience. 

Sealed beam lamps are not capable of producing the sharpness of 
shadows obtained with regular studio lamps. This has been cited as a 
handicap by the cinematographers, although as shown by our Para- 
mount tests, it is not a requisite for pleasing photography. The cam- 
eramen claim the difficulty lies in obtaining an apparent pictorial sep- 
aration of the subject from the background by producing a brightness 

Fig. 5. Lamp housing and the ad- Fig. 6. Multiple exposure photo- 

justable scissor bracket in the re- graph showing movement of the 
tracted position. scissors. 

difference. This decrease in shadow sharpness is due to the increase 
in apparent light source size. For example, the entire area of the 
lamp face of the popular photoflood (RFL-2) is the apparent light 
source. Therefore, when half of the lamp face is covered, no shadow 
is produced on the corresponding side of the set, but only a general 
reduction in total illumination results. Each section of the face of a 
regular studio lamp supplies light to a different area in the set; that 
is, the top portion of the lens face supplies light to only the top area of 
the set and therefore can be conveniently masked near the lamp. 
This permits an actor to be properly illuminated, and then by masking 




Fig. 7. Small set rigged with housings and scissors. 

Fig. 8. Scaffold hangers supported by set walls. 




the outer portion of the lamp face, casts a corresponding shadow close 
to the actor without affecting the amount of illumination on the sub- 

Tests were made to determine in what way and to what extent 
sealed beam lamps deviate from a point source. This was done by 
masking off the face, except for a hole 1 cm in diameter (0.4 in.). The 
1-cm hole was permitted to transmit light at various points located 
from the center out to the edge of the lamp. The plots of a typical 
reflector lamp and a regular studio lamp are shown in Figs. 9 A and 9B. 



20 15 10 5 5 10 I5 e 




Fig. 9A. Distribution of light from small circular areas (1-cm diam.) of lamp face 

of PAR-38 Flood. 

01 2345 6^LAMP FACE 



,-\ x 



N NO. 

/ / 








/ ^ 


A X A 


6 - 





V ..> 



25 20 15 10 5 5 10 15 20 25 

Fig. 9B. Distribution of light from small areas of Baby Junior 
lamp face (full flood position). 


Fig. 10A. Taken with Studio Baby Junior; 
lamp placed 6 ft from subject. 

Note that with the reflector lamp, any point in the field sees the entire 
large source area, while with the regular studio lamp the outer por- 
tions of the lens-face supply light to only the outer areas of the field 
and not the entire field area, thus meeting the requirements for a sharp 
shadow. This effect is pictorially shown in Figs. 10 A and 10B. The 
facial shadows and the shadows from the door molding and door blind 
are comparable using either the studio baby Junior spot or the PAR-38 
flood. But where object and shadow are widely spaced, such as head 
to shadow on the door blind, the baby Junior casts a sharper shadow. 
Note that the shadow break above the door, produced with barndoors 
on housings, is much sharper with the baby Junior than with the PAR- 
38 flood. This indicates that if an improvement in shadow sharpness 




Fig. 10B. Taken with PAR-38 Flood in Research Council housing; 
lamp 6 ft from subject. 

is to be obtained with sealed beam lamps, the face of the lamp must 
be masked to reduce the apparent source size. 

Various types of louvers were tried in front of the lamp face to re- 
duce the apparent source size. In general, the light distribution was 
badly distorted or confined, with loss in light output. An open rec- 
tangular slit kept parallel to the barndoors was found to be the best 
compromise, but the loss of light output makes the use of louvers 

It should be stated that from a practical standpoint the apparent 
source size of a sealed beam lamp cannot be reduced by the use of a 
lens face, such as a Fresnel. This is due to the fact that the filament 


size and focal length of the reflector are the determining factors of 
source size. Light output and size make it impractical to change the 
lamp design. 

Sealed beam lamps should be given serious consideration for use in 
set lighting. Although, as outlined above, sealed beam lamps are 
different in many respects than present studio incandescent lamps, 
they have distinct advantages. 

To summarize, sealed beam lamps supply light at a high efficiency 
and are small and lightweight. For example, a 500-w, PAR-38, 6-hr 
lamp with a Research Council housing weighs 7 Ib and delivers 75% 
of the light of the 2,000-w Studio Junior weighing 39 Ib. Such fea- 
tures make the lamps practical for location work in existing homes and 
buildings where transportation, close shooting quarters and power 
consumption are extremely important considerations. 

Use of sealed beam lamps for stage work will save production time 
if the present technique of "painting" with light is changed. Present 
practice must be modified as it calls for a given number of individual 
light sources, each of which lights an assigned portion of the scene; 
each is adjusted, scrimmed and masked to do its job. Sealed beam 
lamps have no unique value if so applied. However, production 
time can be substantially reduced by the efficient use of sealed beam 
lamps. By efficient use is meant reducing considerably the use of 
light controlling devices, such as gobos, scrims and barndoors. 

In general, our tests, coupled with current studio experience, dem- 
onstrate that sealed beam lamps can be successfully used on both lo- 
cation and stage sets with good photographic quality and a saving in 
production time. 

Color Committee Report 


IT is THE PRACTICE to report to the membership from time to time 
about the work which is being carried on in the different tech- 
nical committees. Due to the excellent support which the chairman 
has received from the members of the Color Committee and its 
subcommittees, it has been possible to make considerable progress on 
some of the projects the committee has undertaken. This report 
outlines briefly the present organization, the scope of its activities, 
what has been done so far, and what is being planned for the im- 
mediate future. 

There is no scarcity of problems which could or should be tackled 
by the Color Committee. It is rather a question of doing first things 
first and giving priority to those problems which are of greatest con- 
cern to the industry. Any suggestions from the members will be 
very much appreciated. 

The eighteen members of the Color Committee represent all major 
organizations actively engaged or involved in the production of color 
motion pictures. 

The objectives of the Color Committee are to make recommenda- 
tions and prepare specifications for the operation, maintenance and 
servicing of color motion picture processes, accessory equipment, 
studio lighting and projection light sources for color, selection of 
studio set colors, color cameras, color motion picture films, and 
general .aspects of color photography. This is a big order and future 
chairmen will not have to be afraid of running out of projects in the 
next five or ten years. 

It was agreed at the first .meeting of the present committee that 
it could best serve the Society by working on a progressive program 
having essentially the following objectives: 

1. To survey the existing information on commercially important 
color motion picture processes and bring that body of informa- 
tion up to date. 

2. To analyze and correlate the technical requirements of color 
motion pictures and evolve recommended practices for the 
guidance of the industry and as forerunners of future efforts for 
standardization . 

3. To disseminate information as soon as it can be organized and 
verified for the edification and assistance of the motion picture 

PRESENTED: April 27, 1950, at the SMPTE Convention in Chicago. 



In order to carry out that program, four subcommittees were or- 
ganized. Lloyd Varden is Chairman of the color process sym- 
posium subcommittee. This subcommittee is working on a review 
of the literature on color motion picture processes which have at- 
tained some measure of commercial success. The chairman is 
arranging with authors or manufacturers to bring publications about 
these processes up to date. New processes not yet adequately de- 
scribed in technical publications are to be included in this symposium. 

In view of the fact that several new processes for color motion 
pictures have been on the verge of commercial introduction, the 
work of this subcommittee has been delayed. It is felt that a thor- 
ough and accurate coverage of the processes, even if it has to be 
somewhat delayed, is preferable to an incomplete or superficial treat- 
ment. Your Chairman would like to add a plea here on behalf of the 
subcommittee chairman for full co-operation of the industry so that 
the information which is needed to make the report worth while will 
be made available in the near future. 

The Color Symposium Report is not intended as a disclosure of 
confidential manufacturers' or consumers' techniques. Its primary 
purpose should be to give a condensed and factual review of the color 
processes available to the industry. 

With a similar intent of providing basic information for the in- 
dustry, a subcommittee on color film sound track characteristics 
has been active, with Lloyd Goldsmith as Chairman. This sub- 
committee has completed its assignment and the report of the com- 
mittee was printed in the March JOURNAL. 

Information about the general principles of color sensitometry has 
been very fragmentary. It has been recognized by the members of 
the Color Committee that there is a rapidly growing need for tech- 
nical information in this field. This is especially true of matters 
relating to the control of new color processes which have become 
available to the industry and which can be processed by the motion 
picture laboratories. 

Sensitometric methods are among the most important tools needed 
to control these processes; therefore, a special subcommittee on 
color sensitometry was organized to study this problem and provide a 
report for the guidance of the industry. The membership of the 
color sensitometry subcommittee has been organized under the chair- 
manship of Carl F. J. Overhage. 

In outlining the scope of the sensitometry subcommittee's activi- 
ties, no attempt was made to provide an immediate solution to 
specific problems which may confront the user of color materials. 
This basic information has to come from the film manufacturer and is 


being supplemented daily by the well-known resourcefulness of the 
laboratory people in the industry. The subcommittee was asked to 
look beyond these immediate requirements and establish the more 
fundamental principles involved in color sensitometry and densitom- 
etry for future guidance. 

Although our present understanding of these processes is still 
incomplete, it was felt that a report presenting the present knowledge 
in this field would be very helpful. It also would lead to the formula- 
tion of basic methods suitable for use throughout the industry, and 
thereby eventually prepare the ground for standardization in the 
field of color sensitometry. 

A report entitled "The Principles of Color Sensitometry" has been 
completed. The report deals with most of the important aspects of 
color sensitometry and contains sections on : 

1. Sensitometric exposure. 

2. The processing of sensitometric tests. 

3. Quantitative evaluation of color images, dealing with the dif- 
ferent types of color densities, such as integral, analytical and 
equivalent neutral density involved in such evaluation. 

4. Densitometer design principles. 

5. Transformation between integral and analytical densities. 

6. Interpretation of sensitometric results. 

7. Statistical aspects of color sensitometry. 

Those who had the opportunity to review this committee's report 
prior to publication unanimously agreed that the subcommittee did a 
very thorough and highly commendable job. The report was sched- 
uled for publication in the June JOURNAL and reprints will also be 
made available in quantity because of the considerable demand that 
already exists for consolidated information on this subject. 

Another subcommittee was established some time ago to investi- 
gate the spectral requirements of light sources and screens for color 
projection. Ronald Bingham is the Chairman of this subcommittee, 
which is working on a report which we hope will be helpful to those 
branches of the industry engaged in furnishing and maintaining pro- 
jection equipment and screens. The main emphasis will be given to 
establishing the theoretically desirable energy distribution of light 
sources for the projection of all presently available color processes. 
In this connection, a study of the dye absorption characteristics of 
commercial two- and three-color processes is being made. At the 
same time, the spectral distribution of various types of light sources 
in use for 16- and 35-mm projection is being reviewed. It is not 
expected that the subcommittee will be able to make definite recom- 
mendations ; however, in an area in which compromises are necessary 


for practical considerations, the subcommittee will attempt to show 
those compromises which will have the least detrimental effect from 
theoretical considerations. 

Another phase of this program will be a review for recommendations 
regarding the spectral reflectance characteristics of various types of 
projection screens for color. 

As far as future plans of the Color Committee are concerned, 
there are at present two subjects on the priority agenda. 

At the last meeting of the parent committee in Hollywood, it was 
suggested that we reopen the subject of phototubes to be used in 
connection with color film sound track reproduction. In 1947, the 
Color Committee prepared a report on the subject of blue-sensitive 
cells for the reproduction of dye tracks. A subcommittee on photo- 
tubes, with Lloyd Goldsmith as Chairman, made a study of the 
blue-sensitive cell and the report of this subcommittee stated that 
there are no important technical objections to the use of blue-sensitive 
cells for sound reproduction. 

It was the consensus of the Color Committee at the time that the 
initiative for the conversion of the blue-sensitive cell would have to 
come from the film manufacturers and the report was temporarily 

As indicated by the review of the subcommittee on color sound 
track characteristics, published in the March JOURNAL, the tendency 
has been toward sulfided sound tracks on multilayer color film mate- 
rials, although it is generally recognized that silver tracks requiring no 
extra processing steps would be preferable. The advent of the lead 
sulfide cell, as yet confined to 16-mm projection equipment primarily, 
makes the whole situation a little more complicated, if not to say, 
slightly confused. 

Little published information is available in regard to the response 
of lead sulfide cells to silver sulfide or dye tracks on color film. The 
Color Committee, in co-operation with the phototube subcommittee 
of the Sound Committee, is now seeking more such information. 

Another subject which the Color Committee intends to take up in 
the near future is the question of color temperature and color tem- 
perature measuring instruments as they apply to color photography. 
It is intended to organize a subcommittee for a study of this subject. 
The objective of this subcommittee will be to review the theoretical 
and practical requirements of color temperature measurements in 
their relationship to color photography. 

The Color Committee will welcome suggestions from the members 
of the Society regarding projects to be tackled by the Committee and 
which are within the scope of the activities of the Color Committee. 

New American Standards 

ON THE FOLLOWING PAGES appear two recently approved 
American Standards for scanning beam uniformity test 
films, and one proposed standard for the sound transmission 
characteristics of theater screens. 

The two test film standards were developed by the joint 
Society and Research Council Committee on Test Films, and 
are based on the old war standard Z52.7-1944. In these new 
standards a departure has been made from past practice in 
preparing test film specifications in that Appendixes have been 
included which indicate the methods of using these films and 
the methods of evaluating the results attained. 

The group responsible for the development of these stand- 
ards believe the Appendixes very desirable because American 
Standards receive wide circulation and are used by many 
people not fully experienced in the field of motion pictures. 

The proposed standard covering the sound transmission 
characteristics of theater screens has been developed 4 by the 
Society's Sound Committee and is also based on a war stand- 
ard Z52.44-1945. However, the transmission characteristics 
specified in this proposal have been met by many types of 
theater screens which have given satisfactory performance in 
theaters for over twenty years. 

On occasion, screens which have excessive transmission loss 
have been installed in theaters. When this has occurred, it 
has been partially offset by raising the gain of theater sound 
system and changing the equalization. In cases where the 
power output of the amplifier is close to the upper limit, such 
a procedure has resulted in excessive distortion. 

Therefore, this proposal is being published for a ninety-day 
trial period. If at the end of that tune no adverse criticism 
has been received, it will be processed as a regular American 





American Standard 

Scanning-Beam Uniformity Test Film for 

1 6-Millimeter Motion Picture Sound Reproducers 

(Laboratory Type) 

Reg. V. S. Pat. Off. 

Z22. 80-1950 

1. Scope and Purpose 

1.1 This standard describes a film which may 
be used for determining the uniformity of 
scanning-beam illumination in 16-mm mo- 
tion picture sound reproducers. The recorded 
sound track shall be suitable for use in labora- 
tories and factories. 

2. Test Film 

2.1 The film shall be a print from an original 
negative. It shall consist of a 1000-cycle, vari- 
able-area recording at full modulation of the 
0.005-inch width and shall be approximately 
sinusoidal. The track shall move uniformly 
0.067 inch from one edge of the scanned 
area to the other as shown in Fig. 1 . 

Fig. 1 

2.2 The position of the sound track relative 
to the ends of the light beam at any instant 
shall be shown by a diagram appearing in the 
picture area, the size and location of which 
is shown in American Standard Location and 
Size of Picture Aperture of 16-Millimeter Mo- 
tion Picture Cameras, Z22.7-1950, or any 
subsequent revision thereof approved by the 
American Standards Association, Incorpo- 

2.3 The scanned area shall comply with the 
American Standard Sound Records and Scan- 
ning Area of 16-Mm Sound Motion Picture 
Prints, Z22.4M946, and the film stock used 
shall be cut and perforated in accordance 
with American Standard Cutting and Perfo- 
rating Dimensions for 16-Mm Sound Motion 
Picture Negative and Positive Raw Stock, 
Z22. 12-1 947, or any subsequent revisions 
thereof approved by the American Standards 
Association, Incorporated. 

2.4 The length of this film shall be approxi- 
mately 34 feet. 

NOTE: A test film in accordance with this standard 
is available from the Motion Picture Research Council 
or the Society of Motion Picture and Television 


(This Appendix is not a part of this American Standard.) 

Before using the above test film it is rec- 
ommended that correct placement of the scan- 
ning beam be determined by means of buzz- 
track test film as specified in American Stand- 
ard Specification for Buzz-Track Test Film for 
16-Mm Motion Picture Sound Reproducers, 
Z22.57-1947, or any subsequent revision 
thereof approved by the American Standards 
Association, Incorporated. 

The uniformity of scanning beam illumina- 
tion may be measured by means of a db meter 

connected to the output of the sound projec- 
tor arrTplifier. The illumination of the scanning 
beam should be adjusted according to the in- 
structions furnished by the manufacturer and 
the variation of the output as registered on 
the db meter should be observed. The illumi- 
nation is considered satisfactorily uniform 
when the output reading as measured by the 
meter is within 1 !/2 db across the entire scan- 
ning slit. 

Approved June 12, 1950, by the American Standards Association, Incorporated 
Sponsor: Society of Motion Picture and Television Engineers 

Copyright, 1950, by American Standards Association, Inc.; reprinted by permission of the copyright holder. 




American Standard 

Scanning-Beam Uniformity Test Film for 

16-Millimeter Motion Picture Sound Reproducers 

(Service Type) 

Rtg. V. S. Pal. Of. 

Z22. 81-1950 

UDC 778.534.4 

1. Scope and Purpose 

1.1 This standard describes a film which may 
be used for determining the uniformity of 
scanning-beam illumination in 16-mm mo- 
tion picture sound reproducers. The recorded 
sound track shall be suitable for use in the 
routine maintenance and servicing of the 

2. Test Film 

2.1 The film shall be a print from an original 
negative. It shall consist of a 1000-cycle, vari- 
able-area recording at full modulation of the 
0.005-inch width and shall be approximately 
sinusoidal. The track shall move uniformly 
0.067 inch from one edge of the scanned area 
to the other as shown in Fig. 1 . 

Fig. 1 

2.2 The position of the sound track relative 
to the ends of the light beam at any instant 
shall be shown by a diagram appearing in 
the picture area, the size and location of 
which is shown in American Standard Loca- 
tion and Size of Picture Aperture of 16-Milli- 
meter Motion Picture Cameras, Z22.7-1950, 
or any subsequent revision thereof approved 
by the American Standards Association, In- 

2.3 The scanned area shall comply with 
American Standard Sound Records and Scan- 
ning Area of 16-Mm Sound Motion Picture 
Prints, Z22.41-1946, and the film stock used 
shall be cut and perforated in accordance 
with American Standard Cutting and Perfo- 
rating Dimensions for 16-Mm Sound Motion 
Picture Negative and Positive Raw Stock, 
Z22.12-1947, or any subsequent revisions 
thereof approved by the American Standards 
Association, Incorporated. 

2.4 The length of this film shall be approxi- 
mately 3'/2 feet. 

NOTE: A test film in accordance with this standard 
is available from the Motion Picture Research Council 
or the Society of Motion Picture and Television 


(This Appendix is not a part of this American Standard.) 

Before using the above test film it is rec- 
ommended that correct placement of the scan- 
ning beam be determined by means of buzz- 
track test-film as specified in American Stand- 
ard Specification for Buzz-Track Test Film for 
16-Mm Motion Picture Sound Reproducers, 
Z22.57-1947, or any subsequent revision 
thereof approved by the American Standards 
Association, Incorporated. 

The uniformity of scanning beam illumi- 
nation may be measured by means of a db 

meter connected to the output of the sound 
projector amplifier. The illumination of the 
scanning beam should be adjusted according 
to the instructions furnished by the manufac- 
turer and the variation of the output as regis- 
tered on the db meter should be observed. 
The illumination is considered satisfactorily 
uniform when the output reading as measured 
by the meter is within 1 Vz db across the en- 
tire scanning slit. 

Approved June 12, 1950, by the American Standards Association, Incorporated 
Sponsor: Society of Motion Picture and Television Engineers 

Univerwl Dccimil Cluiifieitlon 

Copyright, 1950, by American Stgndgrds Association, Inc.; reprinted by permission of the copyright holder, 




Sound Transmission of Theater 
Projection Screens 


1. Sound Transmission Characteristics 

1.1 The sound transmission characteristics of 
theater projection screens shall be such that 
the attenuation at 6000 cycles per second, 
with respect to 1 000 cycles per second, is not 
more than 21/2 db and the attenuation at 
10,000 cycles per second, with respect to 
1000 cycles per second, is not more than 4 db. 

The regularity of response shall be such that 
there is no variation greater than 2 db 
from a smooth curve at any frequency be- 
tween 300 and 10,000 cycles per second. The 
general attenuation at and below 1000 cycles 
per second should not be greater than 1 db. 

2. Method of Measurement 

2.1 The sound transmission of the screen shall 
be measured by means of a loudspeaker, fed 
by an audio oscillator and amplifier, behind 
the screen, and a calibrated microphone, am- 
plifier and output meter in front of the screen. 
The loudspeaker shall be of the type normally 
used in motion picture theaters for the size of 
screen being tested, and shall be placed so 
that no part of the loudspeaker is less than 
2 feet from an edge of the screen with its 
mouth parallel to and separated from the 
screen by the recommended theater installa- 

tion distance of from 4 to 8 inches (center cell 
in the case of a curved front multicellular 
horn). The microphone shall be located 10 to 
12 feet in front of the screen and on the axis 
of the loudspeaker. The sound transmission of 
the screen at any frequency is then the differ- 
ence in the sound level measured with the 
screen in place and with the screen removed. 
2.2 Suitable precautions shall be taken to 
eliminate or minimize the effect of standing 
waves in the test room both in front of and 
behind the screen. 


68th Semiannual Convention 

Society members will meet on October 16 for their 68th Semi- 
annual Convention. The Lake Placid Club, a restful private resort 
in the heart of the Adirondacks, will be the location. Bill Kunz- 
mann, the Society's genial Convention Vice-President invites all 
members to attend, relax in delightful informal surroundings and 
derive maximum value from the many technical papers soon to be 
scheduled. The Papers Committee gives assurance that the pro- 
gram will be well organized, and that it will attempt to provide more 
adequate opportunity for discussion than has been customary for 
conventions in the recent past. 

The growing scope of the Society's interests makes necessary 
adopting this and certain other practices now customary with the 
larger engineering societies. More efficient management of the 
program should result, with greater net benefit to members who at- 
tend technical sessions. Nine technical sessions are being scheduled 
the first, including a Business Meeting, is to begin at 2:00 P.M., 
Monday, October 16. Differing somewhat from the format of pre- 
vious conventions, the Monday evening session will feature the 
introduction of Officers and Governors-Elect and presentation of the 
Society's five major awards. The last session should end at about 
5:00 P.M., Friday, October 20. 

Previous conventions have had their lighter side, and the 68th is no 
exception. The Wednesday night Cocktail Party and Banquet are 
to be followed by a Costume Dance in which all who attend will par- 
ticipate and compete for recognition. Costumes will be as simple or 
as elaborate as the participants may desire. 

Bill Kunzmann has announced appointment of the following Con- 
vention Committee Chairmen and asks that the membership give all 
possible aid to making the 68th Convention the best of all. 

Local Arrangements, E. I. Sponable and W. C. Kunzmann 
Papers Committee 

Vice-Chmn., Montreal, H. S. Walker 
Vice-Chmn., r New York, E. S. Seeley 

Vice-Chmn., Washington, J. E. Aiken 

Chairman, N. L. Simmons 

Vice-Chmn., Chicago, R. T. Van 

Vice-Chmn., Hollywood, L. D. Grig- 


Publicity, Chairman, Harold Desfor 

Registration and Information, E. R. Geib, assisted by P. D. Ries 
Banquet, Hotel Reservations and Transportation, W. C. Kunzmann 
Membership and Subscriptions, Chairman, Lee Jones 

Assisted by A. G. Smith, Atlantic Coast Section Vice-Chairman 
Projection and Public Address Equipment, E. S. Seeley 
Lach'es Reception Committee, Mrs. E. I. Sponable, Hostess 

Co-hostess, Mrs. O. F. Neu 


High-Speed Photography Question Box 

Extensive use of high-speed motion 
pictures and their corollary, time- 
lapse photography as research tools is 
forcing many an engineer to become an 
accomplished photographer, to add to 
his other highly specialized experience. 
He puts the unfamiliar tools of photog- 
raphy to work because he has available 
no other means of securing the informa- 
tion he needs. Although photography 
in its more conventional aspects is 
basically complex, our researcher mul- 
tiplies its complexities many times by 
crowding to the limit nearly every step 
in the process and thereby hands him- 
self, as an amateur, a handful of prob- 
lems that have been stumping the 
experts regularly for years. 

Formal aids, such as high precision 
cameras, accessory optical devices, ex- 
posure measuring instruments and con- 
trol mechanisms, already exist. Infor- 
mation about them has appeared in 
the JOURNAL and in a number of other 
publications, but very little has been 
written about techniques essential to 
good photographic results with cam- 
eras, film and processing under mar- 
ginal operating conditions. To help 
fill this need, the JOURNAL will in the 
future carry periodically a High-Speed 
Photography 'Question Box' wherein 
an exchange of questions and answers 
will provide some measure of con- 
tinuous technique orientation for users 
of high-speed or other scientific ap- 
plications of photography. 

There is little doubt that in many 
cases the problem in one laboratory 
will find a practical solution from the 
experience of another. If you have 
answers to the few questions that ap- 
pear below, please communicate with 
Bill Deacy, Society Staff Engineer at 
the New York office. He will trans- 
mit your solution to the person who 
has the problem and will arrange for 
the answer to be published in a forth- 
coming issue of the JOURNAL. 

f\ 1 High-speed motion pictures 
>C * are required of moving parts 
inside a black bakelite device smaller 
than a dime and about % in. deep. 
An Eastman camera is used operating 
at 4,000 and 8,000 frames/sec with the 
camera 13 in. from the object which is 
illuminated by a pair of No. 2 reflector 
spots placed 6H and 7% m - away. 

A 2-in. lens is operated at //2 with a 
+3 portrait auxiliary attachment. 
Insufficient exposure is obtained using 
Super XX film, and the heat generated 
is such that it alters the performance 
of the device under test. How can the 
illumination be increased and the heat 
removed? What new lighting equip- 
ment or techniques should be used? 

f\ O Motion pictures are being 
> photographed with a 16-mm 
Fastax Camera at 1250 frames/sec, 
using a 6-in. lens, object distance of 8 
ft, Super XX reversal film and two 
750-w reflector spot lamps. The sub- 
ject consists of small parts of a me- 
chanical device, moving at the rate of 
15 to 30 cycles/sec. Specular reflec- 
tions from bright wearing surfaces 
can be controlled with polarizing filters, 
but nearly all other surfaces are ma- 
chined with similar finish so that ad- 
jacent moving parts or areas of inter- 
mittent contact are difficult to dis- 
tinguish in the projected pictures. 
How can these several parts be made to 
stand out more clearly? 

f\ O A manufacturer of air-borne 
Vi ** instruments needs motion 
picture records of vibration effects on 
components of his equipment. The 
instruments under study are small and 
encased, making it necessary to il- 
luminate and photograph through a 
hole in the cover. Vibration frequen- 
cies as high as 800 cycles/sec, with 
total object motion at times of as little 
as .001 in. must be observed. Is it 
possible and practical to study these 
phenomena with high-speed motion 
pictures; and if so, what type of 
camera, lens, exposure meter and what 
frame frequencies will be required? 
Also, what light source should be 
employed in the initial setup? 

f\ A How can a 3 X 5 ft area of a 
>C * dark machine be adequately 
lighted for photography at a frame 
frequency of 3,000? High amperage 
lines are not available. 

f\ C Is special processing for re ver- 
Vc *J* sal film available to users of 
high-speed photography? Longer first 
development would be helpful when 
film is known to be underexposed. 

Engineering Committees 

Magnetic Recording 

The Magnetic Recording Subcommittee under the chairmanship of Glenn 
Dimmick has drawn up and circulated proposed dimensional standards for mag- 
netic sound tracks on 35-, 17^-, and 16- and 8-mm motion picture film. These 
proposals are the result of a three-year attempt to develop standards that will 
meet with universal approval. While isolated opinions hold that some further 
modifications may be necessary, it is the majority opinion that publication and 
wide circulation of the current proposals will help crystallize thinking and lead to 
ear ier agreement. 

Of particular interest in the amateur field is the proposed standard for 8-mm 
magnetic tracks which will permit the addition of sound to 8-mm motion picture 
films. Several manufacturers are now preparing to announce 8-mm projectors 
with magnetic sound reproducers. If these standards, possibly with some modifi- 
cation, can be accepted soon, a tremendous saving for the manufacturers and the 
ultimate users will be realized. Early agreement will also prevent great confusion 
which will result if 8-mm sound films are not interchangeable among the projectors 
of various manufacturers. 

These proposals cannot be scheduled to appear in the JOURNAL in less than sixty 
days, but draft copies are available from Society headquarters. If you wish to re- 
view them prior to publication, write Bill Deacy at headquarters, and he will be 
glad to supply you. 

New Release Print Leader 

Charles Townsend's Subcommittee of the Films for Television Committee is 
working on a proposed first version of a revised type of release print leader. The 
project was undertaken several months ago when the television broadcasters an- 
nounced that the Academy leader Z22.55 does not meet their needs in at least 
three important respects. Precise timing of films, necessary in television broad- 
casting, is not possible. The long series of black frames immediately preceding the 
picture cause excessive flare under conventional television switching procedures 
and also prevent the control engineer from anticipating the normal picture gain 
setting for the picture coming up. Mechanical alignment of projectors and tele- 
vision cameras is particularly critical and should be checked, at least approxi- 
mately, before every picture goes on the air. 

From the outset, it was agreed that any new leader must fill the needs of both 
television and theater interests and should work well on both 35- and 16-mm re- 
lease prints if it were to be unanimously accepted. Otherwise, serious confusion 
would be caused in film laboratories and exchanges by two types of leaders. 

A 35-mm negative of the new proposal will be available very shortly. If you 
desire a trial print, Society headquarters will be glad to supply one in either 16- or 
35-mm width. 

Society Announcements 

New Sustaining Member 

The Society is pleased to welcome Neumade Products Corp. as the most recent 
addition to the rolls of our Sustaining Members. The total now stands at 73, with 
several more now negotiating. The financial support provided furnishes tangible 
evidence of faith in the Society's engineering committee work, and makes it possi- 
ble for more projects to be undertaken and a higher percentage to be completed 
each year. The ambitious publications program, which includes not only the 
JOURNAL but also reports and symposia, also benefits all members. 


The Index for Volume 54 was tacked by two spots of adhesive to the inside back 
cover of the JOURNAL proper in June in an effort to serve both those who have 
wanted the Index bound in the last issue of a volume and those who have as de- 
finitely wanted to receive Indexes as separately bound booklets. 

The spine of the Journal beginning with this issue, shows a rearrangement 
made in an effort to show more clearly the progression and identity of numbers 
within each volume. This has been done chiefly in response to very helpful sug- 
gestions made by Lorin D. Grignon, Development Engineer of Twentieth Cen- 

Briefly Noted 

Radio and Television Law, by Harry P. Warner, 1,095 pp., $35.00, has been re- 
cently published by Matthew Bender & Co., Albany, N.Y. The author is known 
to JOURNAL readers as the coauthor, with J. E. McCoy, of "Theater Television 
Today," in the October, 1949, JOURNAL. Detailed information about the book is 
available in the form of an advertising letter from the publishers. 

Radio file is an index of radio and television articles appearing in the principal tech- 
nical periodicals. This JOURNAL has been added to those previously indexed. 
Radiofile is an index by subject, not by title. It is issued bimonthly and is 
cumulative, listing all material of the current year, so that only the last index 
need be consulted. The year-end Annual is for permanent reference. The yearly 
subscription rate is $1.50. Annuals are: for 1946, $.35; and 1947-49, $.50 each. 
The publisher is Richard H. Dorf, 255 W. 84 St., New York 24. 

"Management Techniques to Match Speed With Efficiency" is an article on com- 
mon sense management by W. Walter Watts in the May, 1950, Dun's Review. 
"Wally" Watts is well known to Society members as Vice-President of RCA and a 
major proponent of theater television. In this article he speaks of coaching an 
industrial organization from within rather than managing from above, as a prac- 
tical and successful way of enabling it to self-adjust to fast changing conditions. 
His is a philosophy that will interest all who are administrators in any industry. 
Dun's Review is 35^ a copy, from Dun and Bradstreet, 290 Broadway, New York 8. 

New Members 

The following have been added to the Society's rolls since the list published last month. 
The designations of grades are the same as those in the 1950 MEMBERSHIP DIRECTORY: 
Honorary (H) Fellow (F) Active (M) Associate (A) Student (S) 

Adams, Stanley H., Factory Representa- Coleman, Theodore T., Motion Picture 

tive, Movie-Mite Corp. Mail: 10609 Producer. Mail: 12610 Brackland 

W. 62 St., Shawnee, Kan. (A) Ave., Cleveland, Ohio. (A) 

Bearman, Alexander A., Engineer, Twen- Cook> Edmund G<> j Ttf Deputy Chief, 

tieth Century-Fox Film Corp Mail: AMC photo Service Center, WP-AFB, 

444 W. 56 St., New York 19. (M) Mftil . Box 11Q7> Wright-Patterson Air 

Bennett, Stanton D., Radio Engineer Force B Dayton, Ohio. (A) 
Radio Station KOMO. Mail: 3437 

36 Ave. W., Seattle 99, Wash. (A) Camels, Victor J Daniels High Speed 

Choudhury, Siraj-ul-Islam, College of the Motion Picture Corp Mail: 395 Barry 

City of New York. Mail: 150-8 Suf- **., Rochester 17, N.Y. (A) 

folk St., New York 2 (Apt. 17). (S) Edwards, Charles N., Photographic Engi- 

Claybourne, J. Philip, Design Engineer, neer, U.S. Naval Photographic Center. 

J A. Maurer, Inc. Mail: 48 Avenel Mail: 2952 Second St., S.E., Washing- 

St., Avenel, N.J. (A) ton, D.C. (M) 


Fallier, Jeptha D., Camerman, Television 
Features. Mail: 31-8433 St., Long 
Island City, N.Y. (M) 
Gippner, Gerald O., Technical-Engineer- 
ing Staff, Movie-Mite Corp. Mail: 
2114 Cleveland Ave., Kansas City, Mo. 

Goldberg, Morris M., Motion Picture 
Photographer, Armed Forces Institute 
of Pathology. Mail: 245 Gallatin St., 
N.W., Washington 11, D.C. (A) 
Gordon, Larry, Producer and Director, 
Television Features, Inc.; General 
Business Films, Inc. Mail: 480 
Lexington Ave., New York 17. (M) 
Halprin, Sol, Executive Director of 
Photography, Twentieth Century-Fox 
Films. Mail: 101 S. Vista St., Los 
Angeles 36, Calif. (M) 
Ham, Richard T., Instructor, Motion 
Picture Photography, The Art Center 
School. Mail: 849 S. Kenmore St., 
Los Angeles 5, Calif. (A) 
Hatcher, George D., Teacher-Television 
Projectionist, Johnstown City Schools 
and WJ AC-TV. Mail: 1184 Agnes 
Ave., Johnstown, Pa. (A) 
Hatcher, Herbert E., Product Designer, 
Bell & Howell Co. Mail: 701 Ridge, 
Evanston, 111. (A) 

Hershman, J. B., President, Radio and 
Television School, Valparaiso Techni- 
cal Institute, Valparaiso, Ind. (A) 
Hessler, Gordon, Film Editor, Films for 
Industry. Mail: 105 Riverside Dr., 
New York 24, N.Y. (A) 
Holmes, Porter, Boston University. 
Mail: 24 Park St., Brockton 48, 
Mass. (S) 

Inderwiesen, Frank H., Radio-Television . 
Engineer, Universal Television School. 
Mail: 1116 W. 40 St., Kansas City 6, 
Mo. (A) 

Leopold, Rudolf, General Mechanical 
Engineer, A. B. Du Mont Laboratory. 
Mail: Demarest Ave., Oakland, N.J. 

Levey, Lawrence, Editor-Publisher. 
Mail: 304 W. 92 St., New York 25, 
N.Y. (A) 

Linden, Michael, Librarian, Motion Pic- 
ture Association of America, Inc. 

Letter to the Editor 

Mail: 168 Washington Park, Brooklyn 
5, N.Y. (M) 

Lockwood, Harold A., Television Engi- 
neer, Farnsworth Television & Radio 
Co. Mail: 3216 Central Dr., Fort 
Wayne, Ind. (A) 

Matheson, Ralph G., President and Gen- 
eral Manager, Matheson Company, 
Inc. Mail: 75 Greaton Rd., West 
Roxbury, Mass. (A) 

Mclntosh, James S., Assistant Director, 
Educational Services, Motion Picture 
Association of America, Inc. Mail: 
7813 Stratford Rd., Bethesda, Md. 

McKnight, Boyd E., Engineer, Minnesota 
Mining & Manufacturing Co. Mail: 
446 N. LaBrea Ave., Los Angeles 36, 
Calif. (A) 

Miller, Thomas H., Manager, Photo- 
graphic Training Dept., Eastman 
Kodak Co., 343 State St., Rochester 
4, N.Y. (A) 

Montes, Ventura, Technical Adviser, 
CMQ Radio Broadcast & TV. Mail: 
Calle A bet. 7th & 9th, Playa Miramar, 
Habana, Cuba (A) 

Phillips, William D., University of South- 
ern California. Mail: Hickory Hill, 
Claremore, Okla. (S) 
Sadkin, Marvin W., Motion Picture 
Laboratory Technician, George W. 
Colburn Laboratory, Inc. Mail: 2925 
W. 56 St., Chicago 29, 111. (A) 
Sandell, Maynard L., Engineer, Eastman 
Kodak Co., 343 State St., Rochester 
4, N.Y. (M) 

Thompson, Orville I., Superintendent, 
DeForest's Training, Inc. Mail: 2533 
N. Ashland Ave., Chicago 14, 111. 

Van Weyenbergh, C., Manager, Western 
Electric Co. (France). Mail: 20, Place 
des Martyrs, Brussels, Belgium. (A) 
Wilson, Willett R., Chief Engineer, Photo- 
lamp Section, Westinghouse Electric 
Corp. Mail: 45 Glenbrook Rd., Mor- 
ris Plains, N.J. (M) 

Young, Robert P., Sales, General Aniline 
& Film Corp., Ansco Division. Mail: 
95 Beekman Ave., North Tarrytown, 
N.Y. (A) 

With reference to Mr. Cummings' letter in the June JOURNAL (p. 766), it is 
noted that the words "the soda ash residue that remains. . ." ignore my previous 
statement ". . . that this film was clean and free of any extraneous matter when it 

There is absolutely no sodium hydroxide, or soda ash as some call it, present on 
the washed and dried film because the material receives a very thorough cleaning, 
both mechanically and by washing, and it is well known that sodium hydroxide is 
very soluble. Furthermore, any minute trace that might be present would cease to 


exist as sodium hydroxide and would be converted into the products of reaction 
between it and the gelatin, and any that might still then be left would be changed 
into sodium carbonate, also very soluble. 

The chance of accidental contamination with sodium hydroxide is quite remote 
because of the method of the washing of the film. 

Mr. Cummings describes the control in nitration as so accurate that there would 
be very little chance of overnitration. 

Without going into too involved a chemical explanation, it is readily conceivable 
that cotton, being a natural product, does not always produce cellulose in exactly 
the same way; differences due to soil, weather, accidental injury to the plant and 
other factors would tend more or less to alter the cellulose, and it is quite possible 
that under these varying conditions some cellulose of the cotton might be suscepti- 
ble to further nitration. 

The writer has seen a blowout occur right at the nitrating spot in a chemical 
plant. The operators thought nothing of it, saying that it was a thing to be ex- 
pected. The nitration kept right on regardless of the blowout because the plant 
was constructed in such a way that it could take care of it. Why did the blowout 
occur if the control is so perfect? 

It is realized that spontaneous combustion due to high nitration is fortunately 
rare, but who knows exactly how rare? The point to stress is that with such a 
substance as cellulose nitrate, the storage conditions should be such as to insulate 
the fire when it does occur, a general point on which both the writer and Mr. 
Cummings agree. 
June 22, 1950 JOSEPH H. SPRAY 

Book Review 

Handbook of Basic Motion-Picture Techniques, by Emil E. Brodbeck 

Published (1950) by Whittlesey House (McGraw-Hill), 330 West 42d St., 
New York 18. i-xiii + 307 pp. text + 3 pp. index. Profusely illus. 6 X 9 in. 
Price $5.95. 

"Right at the outset of this book," says the author right at the outset of his 
preface, "there are a few vital truths which you should know. First is the fact 
that the technique of making motion pictures and the mechanics of making them 
are two different things. Technique is the 'art' and 'skill' of movie making. The 
mechanics of movie making are such things as learning to focus, to expose your 
film correctly, to load and wind your camera." 

To members of SMPTE and readers of the JOURNAL, the mechanics of movie 
making should be an old story. Mr. Brodbeck's first 48 pp:, therefore, may well 
not hold for them anything helpful or revealing. The bulk of his book, however, 
in which in ten major chapters he discusses the "techniques" of movie making 
should be of interest (and perhaps aid) to the practicing technician, especially if he 
makes movies on the side as a personal hobby. 

Mr. Brodbeck's ten chapters take up such subjects as panning, using the tripod, 
shot breakdown, screen direction, matching action, newsreel technique, build-up, 
composition, indoor lighting and applied techniques. Each chapter presents the 
subject in the form of a lesson with text, practice assignments and rules to re- 
member. Mr. Brodbeck's approach to his subject is vigorous and forthright, 
his illustrations practical and informative. On the whole, however, the pictures 
suffer throughout this volume from muddiness of reproduction. 


Home Movies 

New York, N.Y. 


New Products 

Further information concerning the material described below can be 
obtained by writing direct to the manufacturers. As in the case of 
technical papers, publication of these news items does not constitute 
endorsement of the manufacturer's statements nor of his products. 

The Westrex 1035 Magnetic Recording System is a fixed studio or portable loca- 
tion recording system for the use of 35-mm magnetic film at a forward recording 
speed of 90 fpm and a reverse rewind speed of 270 fpm. 

To take full advantage of the inherent signal-to-noise ratio of 35-mm magnetic 
film, a new line of amplifiers having exceptionally low noise has been developed. 
The transmission circuit consists of three amplifiers, two microphone preamplifiers, 
and one main recording amplifier, all basically identical, small in size and with a 
normal flat gain of 70 db from 30 to 10,000 cycles. Their frequency characteris- 
tic may be changed by means of interstage plug-in equalizer units. 

The RA-1467-A Magnetic Recorder is driven by a single-phase, 50- or 60-cycle, 
115-v, synchronous motor, but can also be supplied for operation with a three- 
phase, 50- or 60-cycle synchronous motor, an interlock motor, or a multi-duty 
combination of 220-v three-phase synchronous and 96-v d-c motor. Flutter, or 
"wow," has been reduced to a negligible value. 

Position and dimensions of the recorded magnetic sound track are in accord- 
ance with the proposed Standard 58.301-B of the Academy Research Council. 
The recorder is convertible for use with 16-mm or 17^-mm magnetic film. 

Tight loop threading path is used. The magnetic recording head is mounted 
at the point of optimum scanning and the monitor head is offset for magnetic 
monitoring of the recorded signal. Both the recording and monitor head circuits 
terminate in a single plug and jack connection within the enclosure which can be 
easily reversed to permit using the recording head to reproduce at the point 
of optimum scanning. Consequently the recording machine can also be used 
as a high-quality magnetic re-recording reproducer. 

A driven footage counter adds on the forward run and subtracts on the re- 
verse rewind to keep accurate count of film footage. 

The RA-1484-A Power Control Unit contains a newly developed power supply, 
both line and load regulated, operating from a power source of 115-v, single- 
phase, 50- or 60-cycles. The Control Unit also contains the magnetic bias oscilla- 
tor and the magnetic film monitoring amplifier. 

The entire system is easily transportable and weighs approximately 190 Ib, 
including all interconnecting cables. Further information is available from 
Westrex Corp., Hollywood Div., 6601 Romaine St., Hollywood 38, Calif., or 
Westrex Corp., Ill Eighth Ave., New York 11, N.Y. 


The Line-Up View finder is announced 
by the Hollywood Camera Exchange 
as the first light-weight "Zoom-type" 
viewfinder combining both the 16-mm 
and 35-mm fields, giving the proper 
perspective of the scenes being photo- 
graphed, as well as the area covered by 
any lens. Useful for predetermining 
the proper lens, whether it be a tele- 
photo or an extremely wide-angle lens, 
the calibrations range from 13- to 75- 
mm for 16-mm film and from 25- to 
150-mm for 35-mm film. It is not 
convertible for 8-mm use and it is not 
to be used as an auxiliary lens; nor is it 

designed to be attached to any camera. It measures \Y% in. in diameter by 3^ 
in. long and weighs 2 oz, being carried about the neck on a cord or in one's pocket. 
The price is $15.50, f.o.b. Hollywood. Other information is available from the 
Hollywood Camera Exchange, 1600 Cahuenga Blvd., Hollywood 28, Calif. 

Employment Service 


Wanted : Individual who has had prac- 
tical paid experience in the audio- visual 
field; must have knowledge of film 
storage procedures, circulating and 
maintenance of film, evaluation and 
catalog preparation. Must be able 
to meet the public and to supervise. 
Write: R. E. Herold, 5069 Monte- 
zuma St., Los Angeles 42, Calif. 


Cameraman - Director: Thorough 
knowledge of script-to-screen tech- 
nique. Capable of own script prepa- 
ration and production; 6 yr experience 
free-lance cameraman and producer; 
adept with all types 16-mm photo- 
graphic and editing equipment. Wish 
permanent position with 16-mm indus- 
trial or TV producer; age 27, single, 
free to travel, details readily supplied. 
Robert Deming, 343 S. 13 East, Salt 
Lake City, Utah. 

Producer-Director-Editor: 10 yr with 
major film producers. Thorough knowl- 
edge and experience script-to-screen 
production technique: directing, pho- 

tography, editing, laboratory prob- 
lems, sound recording, 35- and 16-mm, 
b & w and color. Specialist in research 
and prodn. of educational and docu- 
mentary films; small budget commer- 
cial and TV films. Long experience in 
newsreels. Desire greater production 
possibilities, go anywhere. Member 
SMPTE, top refs. E. J. Mauthner, 
P. O. Box 231, Cathedral Sta., New 
York 25. 

Mechanical-Electronic Engineer: B.S. 
degree in Mechanical Engineering; 
extensive design, mfg. experience, 
standard and drive-in theater picture 
and sound equipment; experience as 
engineering assistant to top manage- 
ment exec. corp. in radio TV. Write 
A. Kent Boyd, 3308 Liberty St., 
Austin, Texas. 

On-the-Job G.I. Bill Training; Ambi- 
tious young man to be member of 
camera crew; grad. U.S. Army Signal 
Corps Schl.; experienced with Cine 
Spec., 70DA, Eyemo, Wall and Mit- 
cnell cameras; studied editing, art 
directing and cinematic effects at 
U.S.C.; married, non-drinker, serious; 
man for small studio TV work. P.O. 
Box 524, Alhambra, Calif. 

SMPTE Officers and Committees: The roster of Society Officers 
was published in the May JOURNAL. The Committee Chairmen and 
Members were shown in the April JOURNAL, pp. 515-22; changes in 
this listing will be shown in the September JOURNAL. 


Journal of the Society of 

Motion Picture and Television Engineers 


Characteristics of Motion Picture and Television Screens 


Specifications for Motion Picture Films Intended for Television Trans- 
mission CHARLES L. TOWNSEND 147 

A 100,000,000 Frame Per Second Camera M. SULTANOFP 158 

Flutter Measuring Set FRANK P. HERRNFELD 167 

A Reflex 35-Mm Magazine Motion Picture Camera 


Economy in Small-Scale Motion Picture Lighting ARTHUR L. SMITH 180 

Component Arrangement for a Versatile Television Receiver 

F. N. GILLETTE and J. S. EWING 189 
Designing Engine-Generator Equipment for Motion Picture Locations . 


Laboratory Practice Committee Report JOHN G. STOTT 213 

68th Convention 216 

Board of Governors Meeting 216 

Engineering Committees Activities 217 

Letter to the Editor By Ernest Lindgren 218 

Obituaries 219 

Journals Needed 219 

Central Section Meeting 220 


Film User Year Book, Volume II, edited by Bernard Dolman 

Reviewed by William K. Aughenbaugh 220 
The Organization of Industrial Scientific Research, by C. E. Kenneth Mees 

and John A. Leermakers Reviewed by G. T. Lorance 221 

New Members 221 

New Products 223 

Meetings of Other Societies 224 


Chairman Editor Chairman 

Board of Editors Papers Committee 

Subscription to nonmembers, $12.50 per annum; to members, $6.25 per annum, included in 
their annual membership dues; single copies, $1.50. Order from the Society's General Office. 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and 
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture and Tele- 
vision Engineers, Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General 
and Editorial Office, 342 Madison Ave., New York 17, N.Y. Entered as second-class matter 
January 15, 1930, at the Post Office at Easton, Pa., under the Act of March 3, 1879. 
Copyright, 1950, by the Society of Motion Picture and Television Engineers, Inc. Permission 
to republish material from the JOURNAL must be obtained in writing from the General Office of 
the Society. Copyright under International Copyright Convention and Pan-American Con- 
vention. The Society is not responsible for statements of authors or contributors. 

Society of 

Motion Picture and Television Engineers 

342 MADISON AVENUE, NEW YORK 17, N.Y., TEL. Mu 2-2185 


Earl I. Sponable 
460 W. 54 St. 
New York 19, N.Y. 

Peter Mole 

941 N. Sycamore Ave. 
Hollywood 38, Calif. 


Loren L. Ryder 
5451 Marathon St. 
Hollywood 38, Calif. 

Clyde R. Keith 
120 Broadway 
New York 5, N.Y. 

Robert M. Corbin 
343 State St. 
Rochester 4, N.Y. 

William C. Kunzmann 
Box 6087 
Cleveland 1, Ohio 

Fred T. Bowditch 
Box 6087 
Cleveland 1, Ohio 


Herbert Barnett 
Manville Lane 
Pleasantville, N.Y. 

Kenneth F. Morgan 
6601 Romaine St. 
Los Angeles 38, Calif. 

Norwood L. Simmons 
6706 Santa Monica Blvd. 
Hollywood 38, Calif. 


Frank E. Cahill, Jr. 
321 W. 44 St. 
New York 18, N.Y. 


Charles R. Daily 
113 N. Laurel Ave. 
Los Angeles 36, Calif. 

John P. Livadary 
4034 Cromwell Ave. 
Los Angeles, Calif. 

William B. Lodge 
485 Madison Ave. 
New York 22, N.Y. 

Ralph B. Austrian 
25 W. 54 St. 
New York 19, N.Y. 


Lorin D. Grignon 
1427 Warnall Ave. 
Los Angeles 24, Calif. 

Paul J. Larsen 
6906 S. Bradley Blvd. 
Bradley Hills 
Bethesda, Md. 

William H. Rivers 
342 Madison Ave. 
New York 17, N.Y. 


F. E. Carlson 
Nela Park 
Cleveland 12, Ohio 

George W. Colburn 
164 N. Wacker Dr. 
Chicago 6, 111. 

Edward Schmidt 
304 E. 44 St. 
New York 17, N.Y. 

Malcolm G. Townsley 
7100 McCormick Rd. 
Chicago 45, 111. 

Edward S. Seeley 
161 Sixth Ave. 
New York 13, N.Y. 

R. T. Van Niman 
4501 Washington St. 
Chicago 24, 111. 

Characteristics of Motion Picture 
And Television Screens 



SUMMARY: Two fundamental factors, brightness gain and reflectance or 
transmittance, determine the suitability of a screen material in any particular 
application. High brightness gain, which necessarily implies a narrow view- 
ing angle, may be desirable in one application but not in another. The re- 
flectance or transmittance of the screen material is a measure of the light 

Comparative figures for several commonly used screen materials are pre- 
sented. Both front and rear projection screens are considered. Funda- 
mental photometric concepts are reviewed and laboratory equipment is de- 
scribed. The figures are considered to be accurate to within about 5%, and 
are in reasonable agreement with the few published figures available. 

AT VARIOUS TIMES in ,the past we have needed quantitative infor- 
JL\. mation regarding particular screen materials used in motion 
picture and television projection. A search of the literature on the 
subject revealed very few published figures and very little uniformity 
in the nature of the figures chosen for presentation. The lack of uni- 
formity may be attributed to the fact that there are several systems 
of photometric units in common use, and, further, that certain photo- 
metric terms have been defined differently by various authors. The 
necessity for subsequent interpretation of published data detracts 
from the value of the information. 

To facilitate our investigation, we were obliged to review the neces- 
sary theory, choose a system of units and definitions, and construct 
equipment for measuring the essential screen parameters. The pur- 
pose of this paper is to review the work that has been done and to 
present the results of the measurements which have so far been made. 

The brightness of a screen as viewed by an observer in the audience 
depends not only upon the illumination falling on the screen from the 
projection optical system, but also upon the directional properties 
of the screen. Observers at different positions in the audience may 
see different brightness levels, depending upon the angle from which 
they view the screen. The screen's performance in this respect is 
governed by certain fundamental optical properties of the screen 
material. Before these properties can be discussed, though, optical 
terms which apply alike to all projection screens should be defined. 

PRESENTED: April 25, 1950, at the SMPTE Convention in Chicago. 


132 FEANCE B. BERGER August 

Following the definitions, the basic optical characteristics will be de- 
scribed in as nonmathematical a manner as possible. An exact treat- 
ment requires a mathematical approach, but since the mathematics 
may often obscure the physical concepts under discussion, they are 
relegated to appendixes. 

Characteristics Common to Screens in General 

Of the total incident light that is projected onto a screen, some is 
transmitted through the screen, some is reflected or scattered from 
the screen, and the rest is absorbed by the screen. The fraction of the 
total incident light that passes through the screen is called the trans- 
mission factor or the transmittance of the screen. The fraction which 
is reflected from the screen is called the reflection factor or the reflect- 
ance. The fraction which is neither transmitted nor reflected is called 
the absorptance. These three quantities are often expressed as per- 
centages, their sum being, of course, 100%. For a front projection 
screen a large reflectance is desirable, and the transmittance is gener- 
ally small. For a rear projection screen, large transmittance and 
small reflectance are desirable. The absorptance should be small in 
either case. 

The color of a screen depends upon the spectral composition of the 
light projected onto the screen, and also upon the reflecting or trans- 
mitting properties of the screen material itself. Strictly speaking, 
the transmittance, the reflectance and the absorptance of a screen de- 
pend upon the wavelength of the incident light. For most purposes 
a projection screen should be " white," that is, it should reflect or 
transmit to the same extent light of all visible wavelengths. For the 
purpose of this paper we shall assume that we are dealing with white 
light and with white screens. 

A screen material may be characterized as either specular or diffuse. 
The light transmitted by a sheet of glass, which passes through un- 
changed in its direction of propagation, is an example of regular 
transmission. The light reflected by a mirror leaves at a definite 
angle with relation to the angle of the incident light. Such reflection 
is referred to as specular. For convenience the term specular will be 
used in referring to either regular transmission or specular reflection. 
In contrast to specular effects, a beam of light falling on a blotter is 
reflected from the illuminated spot in all directions. A beam of light 
passing through a sheet of ground glass emerges in all directions. Such 
reflection and transmission are commonly referred to as diffuse. Dif- 
fusely transmitted or diffusely reflected light is referred to as scattered 




Both the transmittance and the reflectance of a material can be 
separated into two parts, the specular and the diffuse. When this 
distinction between specular and diffuse transmittance or reflectance 
is not made, the term total transmittance or total reflectance may be used 
to so indicate. Most materials that are suitable for projection screens 
have small specular coefficients, and one simply refers to the "trans- 
mittance" or "reflectance" of the screen. 

The relative amount, or intensity, of light scattered in the various 
directions is conveniently represented by a polar distribution dia- 
gram. Different screens have different scattering properties and are, 
therefore, represented by different distribution diagrams. A distri- 
bution diagram such as in Fig. 1A characterizes a diffusely transmit- 
ting screen. A screen material having appreciable specular trans- 





Fig. 1. Polar intensity distribution diagrams of rear projection screens: A, 
diffusely transmitting screen; B, screen exhibiting regular (or specular) as well 
as diffuse transmission. 

mission in addition to diffuse transmission is represented by a diagram 
such as that shown in Fig. IB. 

Strictly speaking, polar distribution "diagrams" must be three- 
dimensional diagrams and the distribution "curves" are really sur- 
faces. If the distribution is symmetrical about the normal to the sur- 
face, a simple plane diagram completely describes the directional 
scattering properties of the screen. If the distribution is unsymmet- 
rical, and many practical screens have such unsymmetrical direc- 
tional characteristics, the distribution is commonly represented by two 
plane diagrams; one for the distribution in a vertical plane, the other 
for the distribution in a horizontal plane. 

In the examples cited, it has been tacitly assumed that the maxi- 
mum intensity of the scattered light is observed in the direction nor- 
mal to the screen surface. This may often be the case, but is by no 
means always true. In particular, if the direction of illumination is 




oblique to the screen surface, the maximum illumination is often ob- 
served to be in a direction other than normal to the screen surface. 
Certain possible situations are represented by the diagrams in Fig. 2, 
which pertain to front projection screens. 

It is well to emphasize that an intensity diagram and a brightness 
diagram are not identical. Intensity distribution diagrams have been 
used in the previous examples, but brightness distribution diagrams 
would have served just as well. The brightness in any given direc- 
tion is proportional to the intensity in that direction divided by the 





Fig. 2. Some of the possible types of intensity distributions for front projection 
screens: A, medium brightness gain screen with pattern asymmetrical with re- 
spect to the normal; B, diffusing screen with brightness gain less than unity; 
C, behavior typical of beaded screens; D, high brightness gain screen exhibiting 
marked specular behavior. 

cosine of the angle between that direction and the normal to the 
screen. The two diagrams are therefore related. 

The Brightness Gain of a Screen* 

The directional properties of a screen material can be represented 
to an extent adequate for many projection considerations by stating 
only a single numerical value. This value is the brightness gain. 
In order to define a numerical value for brightness gain, some specific 
directional characteristic must be chosen as a reference. We shall 

* Discussion with Dr. W. W. Lozier following the presentation of this paper 
disclosed that the quantity herein defined as "effective brightness gain" has 
been called "apparent reflectance" by some investigators. 




choose the particular scattering characteristic which defines a Lam- 
bert scatterer as our standard. 

A Lambert scatterer is, by definition, one having a sphere tangent 
to the scattering surface as its three-dimensional intensity distribu- 
tion diagram. It scatters with greatest intensity normal to its sur- 
face. Since it scatters symmetrically about the normal, it is ade- 
quately represented by its plane scattering diagram, which is a circle. 
A Lambert scatterer is often called a perfect diffuser. 

The brightness of a Lambert source, whether an emitter or a scat- 
terer, is independent of the direction from which it is viewed. The 
obliquity decrement in intensity is just compensated by the increase 
in source area corresponding to a constant projected area. A per- 
fectly diffusing surface emitting, transmitting or reflecting N lumens 
per square foot of its area has, by definition, a brightness of N foot- 
Lamberts, which remains the same for all directions of viewing. 

Any screen can be compared with a Lambert screen scattering the 
same number of total lumens per square foot of screen area. Let the 
brightness of the Lambert screen be B' . Let B be the brightness of 
the screen in question when it is viewed in the direction in which it 
has maximum brightness. The brightness gain, G, of the screen can 
be defined as the ratio 

G = - , 

which is equivalent to taking the brightness gain of a Lambert scatter 
as unity. 

The mathematical representation of brightness gain is discussed in 
Appendix I. The above definition of brightness gain is in general 
use. 11 It should be mentioned, however, that another definition is 
sometimes used 7 which gives numerical values just twice as large as 
the values of brightness gain herein defined. 

A screen which has an elongated polar distribution diagram, such 
as that represented in Fig. 1A, has a brightness gain which is greater 
than unity. This type of screen is generally referred to as a direc- 
tional screen. A material having a flattened rather than an elongated 
polar diagram, like that shown in Fig. 2B, has a brightness gain which 
is less than unity; such materials are seldom used for projection 

The brightness of a screen depends not only upon the illumination 
and the brightness gain, but also upon the reflectance or transmittance 
of the screen material. A useful parameter for screen comparison is 

136 FRANCE B. BERGER August 

the effective brightness gain, which includes the effect of reflectance, 
R, or transmittance, T, and is defined as, 

6r e ff = RG (for front projection screens), 
Cr eff = TG (for rear projection screens). 

Choice of Screen Material 

The choice of a screen for use in a given situation depends on how 
the audience is distributed about the screen. The screen should direct 
as much light as possible toward the audience, and as little light as 
possible in other directions. A screen which is "tailored" to the 
audience will make the most efficient use of the available light from 
the screen. 

It is evident that the vertical and the horizontal distribution dia- 
grams of the screen need not be the same. A screen which confines 
the scattered light to the minimum vertical and horizontal angles con- 
sistent with the particular requirements will have maximum usable 
brightness gain. A screen with a lower brightness gain will not 
utilize the available light to the greatest advantage. 

A screen which appears equally bright to all observers within the 
intended region of coverage of the screen and which has zero bright- 
ness to observers situated outside this region cannot be achieved in 
practice. Screen materials can, however, be chosen to approximate 
this condition reasonably well. Parameters which are useful in 
making such a choice are the horizontal and vertical angles of cover- 
age. The brightness gain of a screen is related to these angles of 
coverage, usually defined as the angles between the directions in 
which the screen has half its maximum brightness. 

It is frequently assumed that the light incident on the screen comes 
from a single well-defined direction. This assumption should, how- 
ever, be used with care. In practice, the light incident on any point 
of the screen consists of a cone of rays coming from the projection lens 
aperture and converging at the point on the screen. Further, the 
rays falling on the edges and corners of the screen have a different 
angle of incidence than the rays at the center. In motion picture 
projection, the cone of rays converging at any point on the screen is 
very small, and the rays to opposite corners of the picture make a 
rather small angle with each other. Moreover, low brightness gain 
(wide angle) screens which closely approximate Lambert scatterers 
are generally used. Therefore, in motion picture practice, the as- 
sumption is valid. In television projection, on the other hand, the 




angles involved are quite large and high brightness gain screens are 
generally employed. The range of angles of incidence of the light 
rays at the screen may be comparable to or larger than the angular 
width of the distribution diagram. When the incident convergent 
cone of rays is large, the effective distribution diagrams are broadened 
and the effective brightness gain is lowered, as shown in Fig. 3A. 
When the angle of incidence changes sufficiently over the screen area, 
the distribution diagram differs correspondingly for different regions 
of the screen. This generally will result in nonuniform brightness 
over the screen area, the effect becoming more noticeable at high 
brightness gain figures and at large oblique viewing angles. Curved 




Fig. 3. A, diagrams illustrating the broadening of the intensity diagram that 
results when light is incident over an appreciable range of angles; B, illustrates 
the variation in relative brightness accompanying a variation in angle of incidence 
across the screen. 

screens, 7 auxiliary optical elements such as a Fresnel lens, 11 nonhomo- 
geneous screens, or other innovations may offer advantages in these 

Experimental Equipment and Procedure 

Of the various possible experimental procedures, the following was 
chosen as being the best adapted to measurement of small screen 
samples. A slide projector with a small circular aperture in the 
slide position projects a uniformly illuminated circular spot of light 
onto the screen sample. A visually corrected photronic cell, used as 
a detector, is mounted on an arm which can be rotated horizontally 
about the illuminated spot as a center. The illumination measured 
by the photronic cell is proportional to the candlepower of the ele- 




mentary screen area. A polar plot of a series of measurements at 
different angles gives the horizontal intensity distribution curve. 
The screen sample can be rotated through 90 to obtain the vertical 
pattern. Figure 4 illustrates the apparatus. A known and fixed 
value of incident illumination is maintained by a line voltage regulator 
and a variac. 

Initial calibration is performed with the lamp operating under 
known steady conditions. The total flux projected onto the illumi- 
nated spot is determined by removing the screen sample and allowing 
the light from the projector to fall on a large distant screen. The 
illumination on this large screen area is measured at numerous points 









Fig. 4. Equipment used to determine screen parameters by the intensity method. 

and total flux is determined by numerical integration over the entire 
area. This calibration method gives higher precision than can be ob- 
tained by measurements on the small, highly illuminated spot at the 
screen sample. 

The light output from the projector may change after initial cali- 
bration due to aging of the lamp. To avoid the necessity for recali- 
bration at frequent intervals, a series of measurements was made on a 
carefully prepared magnesium carbonate block. The pattern, bright- 
ness gain and reflectivity of the magnesium carbonate block were de- 
termined as accurately as possible. This block was then used as a 
secondary reference standard. Measurements with reference to the 


magnesium carbonate were found to be satisfactorily reproducible and 
were in agreement with published values. Therefore, as a standardiz- 
ing procedure on subsequent tests, the lamp voltage was adjusted to 
give an arbitrary photronic cell reading with the magnesium carbonate 

Another experimental method was used for measurement of large 
screen areas. In this case, the entire screen is illuminated with a 
projector and the screen brightness is measured directly, using a Mac- 
beth illuminometer or an SEI (Salford Electrical Instrument) expo- 
sure photometer. In order to calculate reflectance and to guard 
against variations in the source, screen illumination is monitored by 
use of a footcandle meter. 

The first method, involving candlepower measurement, is conven- 
ient for laboratory use. Its chief advantage is that only a small screen 
sample is required. The incident illumination can be precisely con- 
trolled. Further, the measurement is based on an electrical meter 
reading and is thus not subject to the human errors which may arise 
in visual photometry. 

The second method, involving direct brightness measurement, can 
be used with large screen samples. The portability of the equipment 
and the nature of the measurements permit use under actual field 
conditions, as in a theater. The method is well suited for measure- 
ments at large angular departure from normal, where candlepower 
falls oft 7 very rapidly but brightness remains relatively constant. 

Some screen samples were measured more than once, and were 
measured by both methods. The consistency of the results leads us 
to believe that the brightness gain values are good to =*= 5% with the 
lower gain figures being somewhat more reliable than the higher 
brightness gain figures. The reflectance values are likewise good to 
about 5%. A series of measurements on magnesium carbonate by 
the candlepower method shows == 2% consistency. 

Experimental Results 

Table I gives the results of laboratory measurement on a number of 
screens and of several miscellaneous materials. Some of the mate- 
rials were measured by the intensity method and others were measured 
by both the intensity and the brightness methods. All data pre- 
sented refer to measurements made with incident illumination normal 
to the screen surface. All of the screens are homogeneous, except 
the ribbed plastic screen with Fresnel lens; measurements on the 
latter pertain to the central region only. Some of the laboratory 
measurements are presented in Figs. 5 and 6. 





Certain optical characteristics of projection screens have been 
dealt with rather extensively. It is hoped that this discussion will 
help the reader to picture more clearly the fairly complex problem 
with which we are dealing and to comprehend the meaning of those 
few parameters which we have considered. 

TABLE I. Characteristics of Representative Screens 

Screen G 



GR or GT 

Miscellaneous Materials 

Perfect screen 





Magnesium carbonate 




Traceolene paper 




Opal glass 




White blotting paper 




Brushed aluminum 




Motion Picture Screens 

Smooth-surface plastic (perforated) 









Nylon cloth 




Metallized directional (perforated) 




Glass cloth 




Commercial Television Screens 

Translucent plastic #1 




Translucent plastic #2 




Diffusing cloth 




Diffusing glass 




Ribbed glass 




Ribbed plastic with Fresnel lens 




Metal beaded 




The experimental apparatus described and the procedure de- 
veloped enable us to measure rapidly the more important optical 
characteristics of a screen material with an accuracy sufficient for 
most purposes. Measurements made on a number of motion picture 
screens, television screens, and other materials of interest, are believed 
to be accurate to within 5%. Our results are in reasonable agree- 
ment with the few published figures we have been able to find. 


The measuring procedure herein described was developed by G. M. 
Rentoumis and the author. Mr. Rentoumis also carried out many 




of the laboratory measurements; B. D. Plakun rendered editorial as- 





Fig. 5. Brightness and intensity patterns for a "perfect" screen (100% re- 
flectance Lambert scatterer), the magnesium carbonate standard, and the smooth- 
surface plastic screen included in Table I. On the brightness diagram, A, the 
radial scale gives foot-Lamberts per foot-candle of illumination; on the intensity 
diagram, B, the radial scale gives candles per TT square feet of screen area per 
foot-candle of illumination. 


Fig. 6. Brightness patterns for a typical beaded screen and for a metallized 
screen. The pattern for a "perfect" screen is shown for comparison. The radial 
scale is graduated in foot-Lamberts per foot-candle of illumination. 

142 FRANCE B. BERGER August 


Derivation of Expressions for Brightness Gain and Reflectivity 

Brightness gain, G, has been defined as the ratio of brightness, B Q , 
of a screen, as observed in the direction in which it has maximum 
brightness, to the brightness BQ of a Lambert screen emitting the 
same total flux per unit area of screen surface. From this definition 
and from basic photometric concepts we shall derive an expression 
for the brightness gain in terms of the brightness distribution curve 
of the screen. Reflectance (or transmittance) will be expressed as a 
function of the gain and of other parameters. 

From basic definitions, we may write, 

dL = Cdu = B da cos du (1) 

where dL is the flux emitted in the elementary solid angle, du, and 
where C is the intensity and B the brightness of the elementary source 
area da when observed in the direction making the angle with the 
normal to da. A consistent set of units must, of course, be employed; 
e.g., L, C, B and da may be expressed in lumens, candles, candles per 
square foot, and square feet respectively. The total flux emitted 
from the elementary source da is then 

L = fdL = fCdu = da fB cos da (2) 

where the integrations are to be carried out over the entire solid angle 
on the observer's side of the plane containing da. 

In general, the brightness of a given source will be a function of di- 
rection which is not necessarily symmetrical about the direction of 
maximum brightness. That is, the brightness, B, must be expressed 
as a function of two variables, say the angles a and ft, where ft is the 
angle between the direction of observation and the horizontal plane, 
and a is the angle between the normal to the vertical element of area 
da and the projection of the direction of observation onto the hori- 
zontal plane. It is convenient to express the brightness, B, in any 
direction as the product of the maximum brightness, B Q , by an angular 
dependence function, g (a, ft), i.e., 

B = B g(a, ft). (3) 

In terms of the co-ordinates a and ft it can be shown that, 

cos = cos a cos ft (4) 

and that the element of solid angle can be expressed as, 

dco = cos ft dad ft. (5) 


Substituting Eq. (3), (4) and (5) into (2), we can now express the 
total flux as, 

7T/2 X/2 

L daJ^B cos 6 du = BtflaJ* jTg(a,ft) cos a cos 2 /3 da dft. (6) 

-7T/2 -7T/2 % 

Now, a Lambert source, by definition, has a brightness independent 
of the direction of observation; i.e., g (a, ft) is a constant equal to 
unity. Setting g (a, ft) equal to unity in Eq. (6) and evaluating the 
resulting simple integral gives for the total flux, L', from a Lambert 
source of area da' and of brightness B f , 

L f = irBo' da'. (7) 

Returning now to our definition of brightness gain, we see that it is 
expressed by the ratio of B Q to J5</, subject to the condition that 
L/da = L'/da'] hence from (6) and (7) we get the desired expression 

G = = . (8) 

Bo' ffg(<*> ft) cos a cos 2 /3 da dft 

We could follow through a similar argument in which attention is 
focused on the intensity distribution rather than the brightness dis- 
tribution. We would then find it convenient to define an intensity 
angular dependence function/ (a, ft) by 

C = <V(, 0) (9) 

analogous to Eq. (3) and our resultant expression for brightness gain 
would be 

G = - (10) 

cos Go y* y* f(a, ft) cos ft da dft 

which is equivalent to Eq. (8) and wherein Go is the angle between the 
normal to da and the direction of maximum brightness. 

The reflectance of a screen is defined as the ratio of the total re- 
flected or scattered flux, L, to the incident flux, L t . The scattered 
flux is given by Eq. (6) and the incident flux may be expressed as 

where E is the illumination on the screen and the integration extends 
over the total area under consideration. If we assume that E is uni- 
form over this area, the integral sign may be dropped and da has the 
meaning previously assigned. Using Eq. (6) and (8) we may write, 

144 FRANCE B. BERGER August 

R = T = W Sf9(<x, cos a cos 2 /3 da dft = ^ (12) 
Li E EG 

where B Q is expressed in candles per unit area and E is in lumens per 
unit area. If the maximum brightness is expressed in foot-Lamberts 
and the screen illumination is expressed in foot-candles, expression 
(12) becomes 

B (ft-L) 1 
- G' 

It follows from this expression that the maximum brightness in foot- 
Lamberts divided by the illumination in foot-candles is numerically 
equal to the effective brightness gain, G eff = RG. Equations (12) 
and (13) give, of course, the transmittance, T, rather than the reflect- 
ance, R, in the case of rear projection screens. 


Treatment of Experimental Data 

The experimental procedures described enable one to obtain a series 
of values of intensity or of brightness measured in different direc- 
tions; i.e., to obtain Co or B and the values of /(a, ft) or of g(a, ft) for 
certain values of a and ft. In order to determine the brightness gain 
and reflectance (or transmittance) from these data it is necessary to 
evaluate the integrals occurring in Eq. (8) or (10). Although it is 
possible to fit the observed data with analytic functional representa- 
tions of g(a, ft) and f(a, ft) it is generally more satisfactory to approxi- 
mate the integrals by numerical methods. 

Experimentally it is convenient and generally adequate to make 
measurements in the horizontal and vertical planes only, i.e., to deter- 
mine /(a, 0) and /(O, ft). Let us use the notation /(a, 0) = H(a) 
for the horizontal intensity pattern, and/(0, ft) = V(ft) for the vertical 
intensity pattern. If 9 is small, very little error is introduced by 
making the approximation, 

f(a, 0) = H(a) F(/3) . (14) 

For simplicity, we shall assume further that the patterns are symmet- 
rical and that 0o = 0, whence from Eq. (10), 

= 4fH(a)dafV(ft) cos ft dft. (15) 

Go o 


Approximating the integrations by summations and expressing the 
angles in degrees we have, 

r/G = - -<Eff(a*)Aa f HE7(0 < )cos 

(57.3) 2 b-i 

where Aa = 90/n, A/3 = 90/m, m and n are any integers, and where 
ctj = A a/2 + j Aa and 0, = A0/2 + i A0. 

In the more general case where the patterns are asymmetrical 
and where Go is not equal to zero, suitable numerical expressions 
superseding (16) can be developed by similar arguments. The 
brightness gain may be computed arithmetically after direct substitu- 
tion of the observed data into Eq. (16) or its counterpart. 

It might be noted that in using apparatus of the type illustrated 
in Fig. 4, the directly measured quantity is the illumination, E d , fall- 
ing on the detector. The intensity C, in candles, of the illuminated 
spot is simply 

C = E d r* (17) 

where E d is in lumens per square foot (foot-candles) and r is the con- 
stant distance from the screen sample to the detector expressed in 
feet. Equation (12) for the reflectance (or transmittance) becomes, 
then, in terms of the directly measured quantities 

R = = = JL (18) 

L t EG EG da 

where E is the illumination on the sample of area da. 

If the direct observations are of brightness rather than of intensity, 
one may compute H and V from the relations 

H(d) = g(a, 0) cos a (19a) 

7(0) = 0(0,0) cos (19b) 

and then use Eq. (16). Reflectance (or transmittance) may be cal- 
culated directly from Eq. (12), necessitating, of course, a measure- 
ment of the screen illumination, E. 

The integral appearing in Eq. (10) may be thought of as the ef- 
fective solid angle occupied by the scattered light. For moderately 
or highly directional screens this solid angle is roughly equal to the 


product of the vertical and the horizontal angles of coverage, A and 
B, expressed in radians. Thus, a rough approximate expression for 
the brightness gain is 

G = . (20) 

AB cos Go 

It is found by trial and error that the approximation is best if A and B 
are defined as the angles between the directions in which the intensi- 
ties have fallen to one-third of their maximum values. 


1. A. C. Hardy and F. H. Perrin, The Principles of Optics, 632 pp. McGraw- 

Hill, New York, 1932. 

2. E. M. Lowry, "Screen brightness and the visual functions," Jour. SMPE, 

vol. 26, pp. 490-504, May 1936 (contains a good bibliography). 

3. R. P. Teele, "Photometry and brightness measurements," Jour. SMPE, 

vol. 26, pp. 554-569, May 1936. 

4. W. F. Little and A. T. Williams, "Resume of methods of determining screen 

brightness and reflectance," Jour. SMPE, vol. 26, pp. 570-577, May 1936. 

5. "Physical concepts: radiant energy and its measurement," Jour. Opt. Soc. 

Amer., vol. 34, pp. 183-218, Apr. 1944. 

6. J. L. Stableford, "Projection screen efficiency," Jour. Brit. Kinematograph 

Soc., vol. 10, no. 1, p. 37, Jan. and Feb. 1947. 

7. W. E. Bradley and E. Traub, "A new television projection system," Elec- 

tronics, vol. 20, p. 84, Sept. 1947. 

8. "British screen brightness standards," Internal. Proj., vol. 22, p. 18, Nov. 


9. "Report of Screen-Brightness Committee," Jour. SMPE, vol. 50, pp. 260-273 

Mar. 1948. 

10. H. G. Boyle and E. B. Doll, "Compact projection television system," 

Electronics, vol. 21, p. 72, Apr. 1948. 

11. R. R. Law and I. G. Maloff, "Projection screens for home television receiv- 

ers," Jour. Opt. Soc. Amer., vol. 38, no. 6, pp. 497-502, June 1948. 

12. H. L. Logan, "Brightness and illumination requirements," Jour. SMPE, 

vol. 51, pp. 1-2, July 1948. 

13. A. G. D. West, "Development of theater television in England," Jour. 

SMPE, vol. 51, pp. 127-168, Aug. 1948. 

Specifications for Motion Picture Films 
Intended for Television Transmission 



SUMMARY: Consideration is given to the major problems encountered in 
television reproduction of motion picture film. Present practices are dis- 
cussed and special requirements of the television system are described. 
The qualities presently desirable in a motion picture film to fit it to that 
system are defined, along with explanatory exposition. An appendix lists 
the specifications for quick reference. 

MANY YEARS of motion picture theater viewing have established 
a reasonably well understood norm for motion picture quality. 
When such motion pictures are transmitted by a television system it 
is presumably desirable to reproduce them with a quality as near that 
norm as the state of the art permits. It is not desirable to demand new 
end-result standards, or to require major changes in production tech- 
niques. It is pointed out, however, that both the film and the televi- 
sion system have inherent technical degradations, which are com- 
pounded in the final result. This discussion is directed toward de- 
fining those properties which a film image should exhibit for television 
use, while still retaining as much as possible of the established motion 
picture characteristics. 

The problem is divided below into considerations of gray-range or 
transfer characteristic, detail rendition or resolution, and scene con- 
tent effects. 


1. Pickup Equipment 

Almost all television stations now use iconoscope tubes for film 
transmission. Under the intensive programs of research directed to- 
ward development of new television pickup devices it is possible that 
superior film transmission systems will become available in the future. 
For the present discussion, however, iconoscope transmission will be 

2. Characteristics of the Television System 

Special projectors are used for television as a means to change from 
the sound film frame rate of 24 per second to the television frame rate 

A CONTRIBUTION: Submitted June 15, 1950. 



of 30 per second. Alternate film frames are scanned three times, as 
against the others, which are scanned twice. This difference in 
handling does not alter the action on the film since the average rate 
of film travel is still the original 24 frames per second. 

The projectors use a very short pulse of light, which can be sup- 
plied by a gap lamp without a shutter, or by an incandescent, or arc 
lamp with a rotating shutter having a very small open period. The 
shortness of the open period causes a severe light loss, so that only 
enough light is obtained for a small picture even though a lamp nor- 
mally intended for large-screen viewing is used. This effect has a large 
influence on the maximum film densities usable, and upon the "noise" 
or graininess of the television system. 

The iconoscope pickup tube has a screen which consists of millions 
of small photocells deposited upon a nonconducting film supported 
by a metallic plate. The photocells "see" the picture for a short 
period, and then are scanned in the dark. They must have good 
memory or "storage" if they are to supply good contrast over the 
whole plate. Also, their small-area sensitivity is influenced by ad- 
jacent area illumination, so that spurious signals are generated, which 
are only distantly related to the original scene content. These signal 
voltages must be balanced out by adding to them other oppositely 
"shaped" voltages, which constitutes the operation of "shading." In 
general, more shading is required when dark scenes are televised, and 
when large black areas are located near a frame edge. The latter 
case usually produces "flare," or a tendency for the black to "bleed 
off" to gray. Flare is an effect which can be reduced by proper film 
density control. 

System "noise level" or electrical graininess* is also important in 
picture quality. It is inevitable that the electronic system should 
produce noise when the signals incoming are low. The amount of 
noise is determined to a great extent by the amount of amplification 
which must be used. A low signal, whether it be due to a dark film, 
or a low sensitivity iconoscope, will require more-than-usual ampli- 
fication and the picture will be noisy. This system "noise grain" is 
important in determining the effects of excess density and graininess 
in the photographic image. 

* Noise level at any point in a transmission system is the ratio of the cir- 
cuit noise at that point to some arbitrary amount of circuit noise chosen as a 
reference. This ratio is usually expressed in decibels above reference noise, ab- 
breviated dbrn, signifying the reading of a circuit noise meter, or in adjusted 
decibels, abbreviated dba, signifying circuit noise meter reading adjusted to 
represent interfering effect under specified conditions. From ASA C42 "American 
Standard Definitions of Electrical Terms." 


Measurements of the transfer characteristic of iconoscopes in- 
dicates that when the above-mentioned effects of storage, shading and 
grain are properly controlled, the tube's output voltage is an almost 
linear function of film density over an appreciable range. This televi- 
sion system gray-range is important to the film manufacturer because 
some films permit a wider range of tones than does the electrical sys- 
tem. Attempting to compress a wide-range picture into a television 
system must result in the appearance of the effects of electrical "toes 
and shoulders/' analogous to characteristics of film emulsions shown 
by the conventional H & D curves. The transfer characteristic en- 
countered looks much like that of ordinary films, in that it is generally 
S-shaped, exhibiting a fairly sharp toe and a more round white-range 
shoulder. There is a reasonably linear central portion of the curve ; 
but, if only that portion is used in reproduction, insufficient output 
voltage is generally obtained, increasing the electrical noise level 
beyond acceptability. Thus it becomes imperative to use as much 
as possible of the iconoscope characteristic, even though some com- 
pression is encountered. 

Problems of compression are somewhat complex. If a completely 
compression-free scene is thrown on the plate of an iconoscope (as in 
the case of live pickup), the characteristic toe and shoulder are not 
excessive. Through long education most observers have come to ex- 
pect and even demand controlled range compression. Such compres- 
sion is valuable in reducing sudden brightness changes between scenes, 
and to permit presentations of scenes which if fully reproduced would 
be painfully bright or so dark as to lose all information. However, if a 
scene is televised which already has a normal amount of compression, 
then that compression and the system compression are cumulative, 
and the result is unsatisfactory. If a film is so made that every bit of 
acceptable toe and shoulder is already present, it would have to be 
reproduced on a toe-and-shoulder-free system. Our present television 
systems do not have that characteristic. 

The effects of a change in exposure or development of a film are apt 
to be obscured when that film is reproduced on a television system 
because video controls permit easy change of contrast and average 
brightness. Some effects which will appear similar to gamma 
changes produced by variations in film processing can be obtained 
by electrical compressions, which can be introduced particularly well 
in the black signal range. Special equipment permits controlled bend- 
ing of the electrical system transfer characteristic to expand the white 
signal range almost at will. Both "amount" and "break point" of 
such bending can be adjusted to suit an individual case. It is, how- 


ever, inadvisable to depend too much upon such compensation, since 
the steep gradients required for large effects cause the electrical sys- 
tem to become "wild" and difficult to operate. In general, films 
should be made to have as little compression in their significant den- 
sities as the state of the art will permit. In this way only small com- 
pensations will be required, and good operational performance can be 

3. Film Characteristics 

There is a real desire on the part of most film producers to make 
their product fit the needs of the television system. Many have asked 
for complete specifications for television films in the hope of obtaining 
inflexibly correct values and procedures. The information below 
will serve to define as far as possible these correct values and pro- 
cedures, but will also indicate why they cannot be inflexibly speci- 

Density range is the most important aspect of the television film. 
To provide good signal-to-noise ratio, the film must transmit as much 
light as possible. To stay within the tone-range of the system no more 
density range should be recorded than is expected to be reproduced. 
The numerical values of these limitations may be stated fairly firmly. 
The usual iconoscope tube can accommodate a density range of ap- 
proximately 1.5. That is, the light energy in the high-lights should not 
exceed 32 times that in the darkest portions to be reproduced. Ac- 
tually, a somewhat greater range can be transmitted, but only with 
excessive compression effects. The position of the 1.5 density range in 
the film characteristic is also important. With present projection 
illumination levels, the minimum significant density should not be 
greater than 0.4 if good signal-to-noise ratio is to be obtained. That 
is, maximum important high-lights should be placed approximately at 
0.4, with some "white peaks" extending below this value only if they 
have little or no importance in the scene, or if no detail is present in 
them. The major reason that lower densities are not valuable lies in 
the fact that most films show compression effects in that range. Print 
characteristic curvatures between the densities of fog level and 0.4 
are usually too great to be tolerated by a system which increases that 
curvature. Of course, there are printing systems which depend upon 
a balance between a long curved negative characteristic and a com- 
plementary print curve. In this case, the above statement may re- 
quire modification, but it is believed that very little printing is done in 
this manner. When "normal" printing is used, best results will be ob- 
tained if the significant densities are placed in the range from 0.4 to 


1.9, with only unimportant areas or areas lacking detail permitted to 
fall outside these limits. In cases where large black areas must appear 
at the bottom of the frame or where black backgrounds are required 
for short periods of time, a further reduction in density of these black 
areas is recommended. A maximum density value of 1.5 in these areas 
is more appropriate if flare and the resulting bleeding of the black are 
to be held to a minimum. Ordinarily such a background will be re- 
produced as black on the television system, since the original intention 
will be obvious to the video operator. 

Many film producers have requested specifications on film stocks 
and development gammas rather than a required range of densities. 
It has not been possible to make any firm statements of this type for 
the following reasons: 

Gamma is the parameter for which most specification requests are 
received. Without any information other than the question, ^'What 
gamma is best for television films?" no answer can be given at all. The 
term "print gamma" usually refers to the value of the slope of the 
density vs. exposure curve for a particular stock developed under some 
particular conditions. Usually it is read at the high-density end of 
the curve. When this is done an excellent measure of the effects of 
development is obtained, but with very little information concerning 
the appearance of the picture. As a tool for processing control, print 
gamma is excellent, but the picture density range is usually not 
"read." Thus two films may be handled so that their IIB densities 
may be plotted as straight lines from values of 1 to 2.5, but exhibit 
entirely different characteristics below that range. They would both 
have the same "gamma," as far as quoting a number is concerned, 
making such a quote an extremely unreliable basis for judging picture 

Again, if a negative is low in contrast, a higher print gamma is re- 
quired than if it were "normal." Both of these conditions can produce 
good pictures, as can the case for a high contrast negative and a low 
gamma print. Obviously, some knowledge of the negative develop- 
ment would be required for print gamma specification. 

"Print-through-gamma" is also an elusive quantity. Curve slopes 
are never read in the actual picture range, especially since evaluation 
of gradients in that range is difficult. The net result of combining 
two gammas is dependent, therefore, on the curve shapes, as well as 
the exact portions of those shapes actually used. Even when a pro- 
ducing company has arrived at standard developments for negatives 
and positives, the assignment of a particular print-through-gamma is 
dangerous because of variations in the original scene contrasts and in 
negati ve'exposure . 


Most film that is good for television use has employed a restricted 
scene brightness range. This does not mean "flat" studio lighting. 
All the accent lighting used so effectively by Hollywood can and 
should be retained. But the ratio of that light to fill-light must be re- 
duced. Again it becomes a problem of fitting the scene into final print 
densities which can be faithfully reproduced. If it is judged that for a 
particular scene the brightness ranges that are high in value are the 
most important, then it may be that large values of back-light and 
high-light can be tolerated, and the densities representing these 
brightnesses must be printed within the range. But if it is judged 
that for another scene the low- value brightnesses are most important, 
then high-lights must be sacrificed, and exposure increased to get the 
print densities down within the specified range. If 100- or 200-to-l 
films are made, only disappointment can result from compressing 
them into a 30-to-l channel. 

Many television films are made outdoors with enormous scene 
brightness ranges, and yet produce excellent results. Some of these 
films actually show a print density range of 3. This case is a good ex- 
ample of the above reasoning. If the wide range film is mostly small 
detail of trees, rocks, etc., the video signal will show no texture in the 
blacks or whites, but none is needed, since these areas are "texture" 
in themselves. As soon as a medium shot or close-up requires detail 
in light or dark areas, those areas must be protected from compression 
by placing them within the specified range. If the large range is 
maintained, faces will become blank white, and dark horses become 
animated charcoal drawings. Judicious use of reflectors or fill-light 
of any kind will reduce the range of most outdoor close-ups to permit 
the adjustment of exposure to produce the densities required. 

From the above it will be seen that many combinations of negative 
gamma and print gamma can be made to yield good pictures, de- 
pending upon the control exercised in original scene brightness range 
and exposure. The final product is a range of densities, which has 
been specified; and the means by which an individual producer ar- 
rives at those values is largely a function of his own operating condi- 


1 . Television System Characteristics 

The television picture delivered to the home viewer is limited in res- 
olution by the bandwidth specified by the Federal Communications 
Commission, and by the performance of the equipment utilized. In 
general, acceptable sharpness is obtainable under normal circum- 


stances. Several studies have been made to determine just how sharp 
a picture can be broadcast in the present television channel. Tests 
with photographic methods whereby an ideal television system can be 
simulated, and the use of actual television equipment of a highly re- 
fined type, have both shown that present system standards can de- 
liver truly excellent definition. That such results are not always at- 
tained can be attributed to the large number of system elements 
which are difficult to control. Some of these are discussed below. 

Under normal circumstances the amplifiers and circuits of the 
television system impose no limitation on the transmission of fine 
picture detail. Pickup tubes, however, can exert a large effect on 
final picture sharpness. Ideally, such a tube should have full video 
voltage output at the highest frequency utilized. That is, the finest 
black-and-white detail to be transmitted should produce as much 
signal voltage as does any larger area. Present-day pickup tubes do 
not completely fulfill this requirement, having a reduced output level 
at the frequencies corresponding to fine detail in the picture. Elec- 
trical equalization is used to compensate for this effect, but this in- 
creases the fine-grain "noise" in the transmitted picture so that large 
amounts of compensation are not desirable. Great care must be ex- 
ercised to see that the pickup tube is supplied with the best possible 
picture, in order that over-all degradation is kept to a minimum, since 
any degradation in the picture will be compounded with degradation 
in the television system. 

Suppose for a moment that an audience will accept without com- 
ment a well-defined maximum loss in picture detail at a certain viewing 
distance. If a television picture having that loss is viewed in that 
manner, acceptable results are obtained. If a film picture having that 
loss is viewed in that manner, acceptable results are also obtained. 
However, if that film, which is acceptable, is viewed over that televi- 
sion system (which in itself is acceptable) seriously degraded and un- 
acceptable pictures will result. Neither picture system in itself is bad, 
but the combination of the systems adds their individual losses, and 
the result is noticeably poor. 

This introduces the idea that each element of a picture transmission 
system must be assigned its appropriate part of the total permissible 
loss. Each such loss must be as small as the state of the art permits. 
Good circuit design has reduced amplifier losses to a negligible value, 
but the enormous complexity of picture tube design and construction 
has not permitted attainment of that degree of perfection in their 
operation. It has thus long been good practice to assign the major 
portion of the total permissible resolution loss to the pickup tube. A 


great deal of research is being devoted to reducing this loss but for the 
present it is well to continue to "pamper" the pickup tube. 

In live-studio practice it is fairly usual for optical systems to de- 
liver to the pickup tube photo-cathode images having limiting res- 
olution in excess of one thousand television lines. Under such cir- 
cumstances very little degradation is contributed by the optical image, 
and the net effective sharpness is that of the picture tube. With 
film, however, projected image resolution rarely reaches such a high 
value, and the net effective sharpness is below that of the pickup tube 
alone. It is interesting to note that live-studio pictures are noticeably 
degraded when the optical resolving power drops below 800 television 

Further complicating the resolution problem is the electrical grain 
or "noise" inevitable in present systems. If pictures having small 
"signal" content (low density range) are fed to such a system, am- 
plifier gain must be raised beyond normal limits to regain normal 
operating levels. This increases the effect of noise, masking the fine 
detail in much the same manner as does the grain in a poorly made 

Kinescope picture viewing tubes also are pertinent in a discussion of 
resolution. Good tubes having fine spots are available, but generally 
some loss should be allowed for this device. An effect called "bloom- 
ing" is particularly important in film reproduction. Whenever an 
excessively wide gray-range is fed into the television system, very 
bright white areas are likely to produce high signals which are well 
above the general "tone" of the scene. In order to reproduce the 
lower signals properly, the voltages must be amplified more than 
usual before being fed to the picture tube. In this case the bright 
white signals are too high in level for normal operation and those 
areas blur, losing line structure and picture texture. A reasonable 
balance between "whites" and "blacks" is desirable for maximum 

2. Film Capabilities 

Having established the resolution needs of the television system, it 
becomes possible to define the performance required of film systems 
designed for its use. Again, some portion of the total permissible res- 
olution loss must be assigned to the photographic medium. But 
every effort must be made to match the live pickup sharpness, which 
means that very little loss can be so assigned. Photography is an old, 
established craft capable of excellent image sharpness, so it seems 


reasonable that stringent requirements should be placed upon it, 
leaving more leeway for the infant television art. 

Quite often it is said that film has such excellent resolution that 
there cannot be any problem in its television use. Published values 
of limiting resolution for many films seem to confirm this, but a closer 
investigation indicates differently. First, it must be remembered 
that the film resolving-power ratings are for "cutoff" conditions. 
That is, they state the highest value at which any line structure can 
be seen. This, of course, is at a very low contrast far too low to be of 
any value to the television system. "Contrast" is "modulation" in 
the electrical system, and it is possible to plot the response of film in 
much the same manner as an electrical system. When this is done, it 
is discovered that films have no "flat bandwidth." That is, their 
contrast falls off as the size of the elements to be resolved decreases. 
If the total photographic system is allowed about 10% loss in contrast 
at the maximum television resolution, it is found that its limiting res- 
olution value must be well above the television cutoff. As a result of 
this, the best 16-mm films will be found to be barely good enough. 
As a matter of practical fact, it is exceedingly difficult to realize a res- 
olution of 400 television lines with a high value of contrast in an 
ordinary 16-mm release print. Such a print includes degradations due 
to all the elements of the photographic system, including the effects 
of printing. For the present, only the very best products and tech- 
niques can be combined to produce a 16-mm print which will not 
seriously limit the results obtainable through the television channel. 
Whenever feasible, 35-mm film should be used, and in this case also, 
the best methods should be followed. Unfortunately, not all televi- 
sion stations are equipped to transmit 35-mm film, but if original 
shooting is done in that size, good quality can be expected from most 
of the larger stations, and the rest can be served by reduction-printed 
16-mm versions. 


1. Reproduced Area 

Reference is made to the Television Test Film of the Society of 
Motion Picture and Television Engineers. The projector alignment 
section of that film includes an implied standard definition of the area 
to be scanned. Many stations now have copies of this film, and it is 
believed that following its directions will lead to satisfactory results. 
Approximately 1^% of the standard projector frame is cut off in 
scanning at the top and bottom of the frame, and the sides are cut by 


an amount required to maintain the 3X4 aspect ratio specified by 
law. The side losses are not the same for 35-mm film as for 16-mm 
film, due to the different film frame aspect ratios. 

Alignment chart sections from the 35-mm and 16-mm versions of 
the test film can be purchased separately, thereby reducing costs. 
Frequent reference to these charts, along with the instruction book 
accompanying them, is recommended. 

Also included in the above chart is a rectangle enclosing approxi- 
mately 65% of the frame area. The lines forming it are placed so as 
to produce a 10% border within the televised frame. The area within 
the rectangle is believed to be reasonably well reproduced on the home 
receiver even when scanning is poorly adjusted and centering is badly 
set. Important information should be kept within this area, es- 
pecially commercial copy titles or trade-marks. 

2. "Busy" Scenes 

Care should be taken to insure that a scene being photographed 
does not have a high-contrast background that will detract from 
foreground action when the picture is viewed on a small screen. 
Simplicity seems to be required in backgrounds for television more 
than for theater use, where images are not "crowded" by the frame 

3. Shot Sizes 

Television has long made good use of close-ups and medium shots. 
Small screen sizes are not the only reason for this. The resolution 
needed for a close-up is less than for a long shot, merely on the basis 
that less fine detail is needed to carry the intended information. 
Thus, receiving sets which are mis-tuned or are out of focus will 
reproduce close-ups when long shots will be hopelessly blurred. 

The above is not intended to eliminate long shots. Establishing 
locale and impressions of size are as important as in the theater, but 
important details of wide-angle shots should be pointed up with 
clever accent lighting and reduction in unimportant competing de- 

4. Scene Tone Balance 

Some refinements in smoothness of reproduction can be obtained 
when large black-and-white areas are needed, if they are used with 
care. Half-black, half-white pictures, with the dividing line running 
horizontally, usually require relatively large shading corrections. 
This is particularly true if the lower half of the picture is black. Sea- 


scapes or any sky and land scene can fall into this category if no large 
foreground objects are available to break up the pattern. Also, sudden 
large changes in scene brightness should be avoided, as they place 
severe requirements on both transmitter and receiver "d-c insertion" 
performance. Smooth changes or small steps are usually reproduced 
without trouble. Cutting between shots of an object which have 
radically different background brightnesses can cause the object itself 
to appear to change tone, becoming darker with the light background, 
and lighter with the dark background. Avoid if possible the use of 
full daylight shooting of night scenes when the required effect is pro- 
duced by purely photographic means. "Blacks" look severely com- 
pressed in that case, and video operators tend to raise their brightness 
control to bring out what may be there, but is not. If possible, al- 
ways include some full-level high-light to define the "white signal" 
limit. Usually this can be done without harming the scene mood for 
direct projection and will greatly aid in television transmission. 


1. Density. Normal contrast range, 1.5; minimum density, 0.4; and 

maximum density, 1.9. 

2. Gamma. No exact statement possible, but generally the above will 

require that gamma be somewhat lower than usual in films in- 
tended for theater use. 

3. Resolution. Limiting value, minimum, 800 television lines. 

4. Scene content 

Follow : SMPTE frame-size specifications in the Television Test 

Avoid : Sudden large brightness changes, large black areas near 
frame edges, "busy" backgrounds, too great gray range, arti- 
ficial night shots. 


1. Otto Schade, "The electro-optical characteristics of television signals," RCA 

Rev., (Publication #ST353, Tube Dept., Harrison, N.J.), 1948. 

2. R. B. Janes, R. E. Johnson and R. S. Moore, "Development and performance 

of television camera tubes," RCA Rev., vol. 10, no. 2, pp. 191-233, June 1949. 

3. D. R. White (Chairman), "Films in television," Jour. SMPE, vol. 52, pp. 363- 

379, Apr. 1949. 

4. G. D. Gudebrod, "Television-film requirements," Jour. SMPE, vol. 53, pp. 

117-119, Aug. 1949. 

5. A. J. Miller, "Motion picture laboratory practice for television," Jour. SMPE, 

vol. 53, pp. 112-113, Aug. 1949. 

6. The Use of Motion Picture Films in Television, 56 pp., Eastman Kodak Co., 

Motion Picture Film Dept., 343 State St., Rochester, N. Y., 1949. 
See also: "Television test film," Jour. SMPTE, vol. 54, pp. 209-218, Feb. 1950. 

A 100,000,000 Frame Per Second 



SUMMARY: Shock waves close to the edge of explosive charges have been 
successfully photographed at rates exceeding 100,000,000 frames/sec. 
These ultra high framing rates are obtained with a multi-slit focal plane 
shutter which is transported optically across the film plane by a rotating 
mirror. Linear shutter speeds up to 3,000 meters/sec are easily obtained, 
and the resulting framing rates with the proper selection of slit widths can be 
varied from 10 5 to 10 9 frames/sec. Each individual frame is composed of a 
series of lines, and the degree of "discontinuity" across each frame is pro- 
portional to the total number of frames. 

THE EXPERIMENTAL STUDIES of the shock and detonation which 
accompany explosive reactions have been hampered by the lack 
of ultra high-speed instrumentation. Short duration optical studies 
are particularly required for the investigation of self-luminous detona- 
tion and shock waves. 

The velocity of these transients averages about 8 mm per micro- 
second; therefore, usable photographic exposures of these transients 
must not exceed 10 ~ 7 sec. Kerr cell shutters 1 have been used to ob- 
tain a single or a few successive short duration exposures, while 
multi-lens cameras 2 have produced continuous short duration ex- 
posures, but at rates which are not adequately high. The O'Brien- 
Milne camera, 3 which is rated at 10,000,000 frames/sec, but which 
displays poor resolving power, could not be obtained commercially, 
and its precise optical system made it impractical to build locally. 

A motion picture camera which employs simple optical and me- 
chanical systems to obtain up to 300 successive 4 X 4 in. frames at rates 
which can be varied from 10 5 to 10 9 frames/sec, and which exhibits 
satisfactory resolving power, is described in this paper. 


The standard variable slit focal plane shutter in common use ex- 
poses a time-space record as it travels across the film plane. Al- 
though the slit moves slowly across the film, the exposure time can be 
made extremely short by reducing the width of the slit. 

PRESENTED: April 26, 1950, at the SMPTE Convention in Chicago. 

100,000,000 FRAMES PER SECOND 


The framing grid is a focal plane shutter with a series of parallel 
slits placed at regular intervals across it. This shutter, therefore, is 
required to move only the distance between two successive slits to 
expose the entire film. To understand how this grid records suc- 
cessive frames, consider a series of optically clear slits .0005 in. wide, 
cut at .015-in. intervals across a 4 X 4 in. optically opaque plate. 
If this grid is held in a fixed position on a 4 X 5 in. photographic plate, 

Fig. 1. Single still photograph of spherical charge in firing 
position, taken through .0005-in. slit, .015-in. space grid. 

a single exposure made through it would consist of a set of parallel 
lines which occupy only ^Q of the total picture area with an over-all 
dimension of 4 X 4 in. A sample exposure of this type is shown in 
Fig. 1. By moving the grid across the film perpendicularly to the 
slits for a distance of .0005 in., and exposing a second still picture in 
this new position, a second series of lines lying alongside the first set 
and again occupying only ^ of the total picture area will be pro- 




duced. Thirty such single pictures will result from only .015-in. move- 
ment of the grid, and will expose the entire film area. To the casual 
observer the resulting picture will be an indistinguishable jumble. 
However, by proper positioning of the grid, any one of the 30 ex- 
posures can be studied separately. This type of grid framing has 
been used for years in animated greeting cards and photographic ad- 

If the photographic object is moving, and if the grid is moved at a 
uniform rate across the film for a distance of .015 in., the resulting 
picture can be viewed through the grid as 30 separate exposures, one 
at a time, or, by viewing through the grid moving at any uniform 







Fig. 2. Synchronizing circuit for ultra high-speed camera. 

speed, flickerless motion pictures will be observed. This adaptation 
of grid framing has been described in papers delivered by Dr. Fordyce 
Tuttle of the Eastman Kodak Co. 4 

At this laboratory we have been successful in combining a stationary 
framing grid with a rotating mirror to obtain framing rates in excess 
of 10 8 frames/sec. 


The optimum slit width for the multi-slit focal plane shutter ap- 
pears to be of the order of .0001 in. A shutter with .0001-in. slits 
is required to move 10,000 in./sec to produce 10 8 frames/sec. It is 

1950 100,000,000 FRAMES PER SECOND 161 

impractical to accelerate to, maintain, and decelerate from, such 
high velocities with a linearly moving shutter. A rotating focal 
plane shutter, on the other hand, has the double disadvantage of re- 
quiring tapered radial slits and the combination of a large diameter 
and high rotational velocity. A method for moving the image of the 
shutter across the film plane by reflection from a rotating mirror was 
obviously a simple solution to this problem. The rotating mirror 
optical system in a Bowen RC-3 Rotating Mirror Camera, 5 although 
not adequate for this application, was available at this installation, 
and it was modified to take a 4 X 4 in. multi-slit framing grid and a 
4 X 5 in. camera back. 

A first lens is used to image the event on the grid. The combine 
event-grid is then focused by a second lens, whose image, after re- 
flection from the rotating mirror, falls on the film plane as shown in 



Fig. 3. Optical system of ultra high-speed camera. 

Fig. 3. The photograph formed by the reflected grid differs from that 
formed through a grid moving across the film plane in that the latter 
records a varying-time-varying-space image, while the former, by 
virtue of the subject's fixed position with respect to the grid, records 
a fixed-space-varying-time record which is particularly suited to the 
studies of detonation and shock waves. 


The continuous motion of the image across the film does not form 
the well-defined frame of the intermittent or the rotating prism type of 
cameras. An infinite number of viewing positions of the framing 
grid are possible, leaving the framing rate undetermined. However, 
since the exposure time of any increment of film is equal to the time it 
takes a slit to move its own width, the reciprocal of this time is taken 
as the framing rate. Thus, each consecutive frame is viewed by mov- 
ing the grid one slit width per frame. 






1950 100,000,000 FRAMES PER SECOND 163 

The rotational speed of the mirror being used can be varied up to 
500 rps; with a 20-in. optical arm the maximum image speed is 
.122 in. per microsecond. The exposure time with a .0005-in. slit at 
the full rotational speed of the mirror is 4 X 10 ~ 9 sec, or 2.5 X 10 8 
frames/sec. With a .0001-in. slit the framing rate is 1.25 X 10 9 
frames/sec. With suitable optics, it is believed that sufficient light 
will be available from detonation and shock phenomena to take 
pictures at the rate of one billion frames per second. 


The resolving power, or more exactly, the measure of the "dis- 
continuity" across the picture because of the nature of its line struc- 
ture, expressed in lines per inch, is equal to the number of slits per 
inch. The number of slits per inch is determined by the slit width 
and the number of frames required. That is, with a .0005-in. slit if 
30 frames are desired, the space between the slits must be 30 times 
.0005 in. or .015 in. Such a grid will have a "resolving power" 
across the slits of I/. 0005 + .015 or about 65 lines/in. With a 
.0001-in. slit and the same resolving power, 150 consecutive frames are 

The total exposure time, with no double exposures, is the product 
of the exposure time per frame and the total number of frames without 
dpuble exposures. With 30 frames taken at 100,000,000/sec, the 
total exposure time is 3 X 10 ~ 7 seconds. All of the pictures taken up 
to the present time have been of self-luminous shock waves. Double 
exposures are prevented by quenching the shock wave in an atmos- 
phere of propane at a predetermined time. However, for investigat- 
ing shock velocities, it has been found convenient to get multiple ex- 
posures (Figs. 4a and 4b), which permits the measurement of distance 
as a function of time on each individual frame. 


The components of this camera, shown in Figs. 5-7, are described 

1. First Imaging Lens. One of a number of high-quality large 
aperture photographic objectives which are available is used depend- 
ing on the size of the charge and the magnification of image desired 
at the framing grid. All of these lenses are long focal lengths (12 to 
40 in.) as required by the physical setup in the blast chamber (Fig. 2) . 

2. Framing Grid. The grids most easily obtained and reasonably 
priced are made on opaque coated optical glass 4 X 4 X M m - The 



Fig. 5. Side view of camera. 


Fig. 6. Three-quarter front view of camera showing grid in position. 


100,000,000 FRAMES PER SECOND 


.0005- and .0001-in. slits are cut with corresponding flat diamond 
points on a dividing engine. 

3. Second Imaging Lens. This lens has a focal length of 360 mm. 
With the 1 : 1 magnification from the grid to the film plane the mirror 
can be placed to give an effective optical lever of about 20 in. Several 
photographic objectives with apertures of about //4.5 have been 
tested. A high-quality process lens corrected for a flat field appears 
to resolve the slits most accurately. 

4- Rotating Mirror. The 1-in. square face octagonal mirror in the 
Bowen rotating mirror camera is the aperture stop of the system. A 
2-in. square flat mirror designed to rotate around its face will not 



Fig. 7. Top view of camera showing optical system. 

only take advantage of all of the light from the second lens but will 
eliminate the complicated form of the focal plane. 

5. Film Plane. A standard 4 X 5 in. camera back is being used as 

6. Synchronizing Circuit. A hole in the wheel which drives the 
mirror passes a beam of light to a phototube which in turn fires a 
thyratron. This thyratron, operating at 400 v, then fires the charge 
directly or is used to trigger a timing circuit depending on the time 
requirements for the particular charge. 


Kodak Tri-X panchromatic plates have produced images of good 
density at exposure times of 10 ~ 8 sec. These plates are processed 


in Kodak D19 developer for 20 min at 60 F. Base fog "latensifica- 
tion" of these plates has been used to obtain good image densities with 
developing times as short as 8 min at 65 F. 


Since the magnification from the grid to the film plane is 1:1, the 
negatives or contact printed positives can be viewed by placing the 
grid directly over them. Motion pictures can be observed by mov- 
ing the photographic plate across the grid and frame-by-frame view- 
ing is accomplished with a micrometer feed. Measurements are 
made at a magnification of 10 with a Bausch & Lomb contour pro- 
jector. The sample reproductions in this report were made by en- 
larging through the grid properly positioned on a positive plate. 


The fundamental features of the design described in this report in- 
dicate limitless possibilities for obtaining ultra high-speed photo- 
graphs of self-luminous phenomena. Probably, with the use of ex- 
plosive-type flash bombs, or with Edgerton-type lights, nonluminous 
subjects can also be photographed at rates exceeding 10 8 frames/sec. 


1. A. M. Zarem, Marshall and Poole, "An electro-optical shutter for photographic 

purposes," NAVORD Report 1016, May 14, 1948. 

2. J. Stanton, "The Bowen 76 lens camera," (a preliminary description). 

3. Brian O'Brien and Gordon Milne, "Motion picture photography at ten million 

frames per second," Jour. SMPE, vol. 52, pp. 30-40, Jan. 1949. 

4. Fordyce E. Tuttle, "High-speed motion pictures by multiple-aperture focal- 

plane scanners," Jour. SMPE, vol. 53, pp. 451-461, Nov. 1949. 

, "Improvements in high-speed motion pictures by multiple-aperture 
focal-plane scanners," Jour. SMPE, vol. 53, pp. 462-468, Nov. 1949. 
(The above also appear in High-Speed Photography, vol. 2, published by the 
Society of Motion Picture Engineers in 1949.) 

5. I. S. Bowen, "The CIT Rotating Mirror Camera (Mod. 2)," April 27, 1945. 

Flutter Measuring Set 


SUMMARY: The Flutter Measuring Set was built to measure the low per- 
centage flutter of present-day recording and reproducing equipment. The 
set conforms to the "Proposed Standard Specifications for Flutter or Wow as 
Related to Sound Records," as outlined in the Society's Journal for August, 
1947, pp. 147-159. 

IN KEEPING with the proposed specifications cited above, the instru- 
ment provides means for measuring percentage of flutter and drift. 
It will measure flutter at the nominal frequency of 3000 == 200 cycles/ 
sec. The input required for operation may be either +8.0 or 
52.0 dbm. At these inputs, amplitude variations of =*= 10 db will 
not change the measurement by more than 2%. The percentage 
flutter meter is calibrated in percent for 0.1, 0.3 and 1.0% full scale 
deflection. It indicates either true rms or average values by switch- 
ing into the meter circuit either a thermocouple or selenium rectifier. 
The percent drift meter is calibrated for 0.1, 0.3 and 1.0% drift. 
The drift meter is also used as an indicator to tune the Flutter Measur- 
ing Set to the incoming signal frequency. The set is capable of read- 
ing rates of flutter from to 200 cycles/sec. Networks are provided 
to read the following rate bands on the percent flutter meter: 2 to 
200, 2 to 20, 20 to 200, and 96 cycles/sec. The drift meter is used to 
read to 2 cycles/sec flutter rates. Outputs for a rapid recording 
oscillograph and an oscilloscope are provided. 


The flutter set consists of the following components : 

(a) A pre-amplifier. 

(b) A modulator-oscillator which converts the incoming 3000-cycle 
signal to 1000 cycles. 

(c) A limiting amplifier. 

(d) A frequency discriminating network which converts frequency- 
modulation into amplitude-modulation. 

(e) An amplifier to increase the amplitude of the modulation signal. 

(f) A selective network to break down the rate of the flutter into 
. several bands. 

(g) An indicating device. 

PRESENTED: April 24, 1950, at the SMPTE CONVENTION in Chicago. 





The block diagram of the flutter set is shown in Fig. 1. With the 
input switch in the 52.0-db position, the flutter set is designed to 
work from a nominal 500- to 600-ohm impedance. The input is un- 
grounded. The wiring of the input plug is so arranged that the flutter 
measuring set can be grounded on one side, or it can be used in a 
balanced circuit by grounding the centertap of tl^e input transformer. 

The pre-amplifier consists of a nominal 500- to 60,000-ohm input 
transformer and two high-gain pentode tubes. The voltage gain 
from the 500-ohm input to the output of the second stage is adjusted 
to 60 db. Negative feedback is used to adjust the gain and to reduce 
the output impedance of the amplifier. No attempt has been made 















r ^ 












2-20 C\ 20-200 
2-200^ ' 96 


Fig. 1. Block diagram of flutter set. 

to use the negative feedback for improving the frequency character- 
istic as the amplifier handles only a single frequency, namely, 3000 

The input switch either connects the signal to the input transformer 
of the pre-amplifier giving the instrument a sensitivity of 60.0 db 
relative to 6 milliwatts or connects the input directly to the volume 
control preceding the modulator tube, giving the instrument a sensi- 
tivity of 0.0 db relative to 6 milliwatts. With the input switch in the 
0.0-db position, one side of the circuit is grounded and the input 
looks like a true 4,000-ohm resistance. 

In either position of the input switch an 85 volt polarizing voltage 
for a photoelectric cell appears on the input plug. This is done to 



facilitate the use of the instrument, in the high-gain position, directly 
from a photoelectric cell. 

The circuits for the modulator, oscillator and low-pass filter are 
shown in Fig. 2. The plate of the modulator tube is fed through the 
plate load resistance of the oscillator V2. The oscillator is of 
the electron-coupled type. Variations in the plate impedance of the 
modulator tube reflect very little into the oscillator section. The 
input to the modulator tube can be varied over =*=10 db from its 
nominal " calibration" input without changing the frequency of the 
oscillator more than 1 cycle at a mean frequency of 4,000 cycles. 

Figure 2. 

This means that the instrument is capable of handling a signal with 
large amplitude changes without giving erroneous flutter readings. 

The oscillator tunes from less than 3,800 to over 4,200 cycles. The 
discriminating network works at a frequency of 1,000 cycles; there- 
fore, any input signal from 2,800 to 3,200 can be handled by the set. 

The modulator tube is cathode loaded by RI. RI is chosen of such 
value that it is equal to the output impedance of the modulator tube : 

RI = Load Resistance 

n = Amplification Factor of Tube 

R p = Dynamic Plate Impedance of Tube 

The image impedance of the low-pass filter following the modulator 
is equal to Ri t and its cutoff frequency lies at 1300 cycles. The low- 




pass filter consists of two full constant-K sections, LI and Ci. It will 
attenuate the signal and the oscillator tone by about 54 db, but will 
pass the beat, that is, the oscillator minus the signal frequency, unim- 

The output from the low-pass filter is fed directly into a three-stage 
limiting amplifier. 

A duo-diode is located between the second and third tube of the 
limiting amplifier. It is connected in such a manner that it will clip 
both the negative and positive peaks of the signal. 

The clipping starts at about 16 db below the normal input level of 
the flutter set. Clipping and the inherent insensitivity of the dis- 
criminating network to amplitude variation will make the output 
readings of the instrument virtually independent of variations in input 

Figure 3. 

voltage. The last stage of the limiting amplifier feeds into the dis- 
criminating network directly. Figure 3 shows the discriminator and 
allied circuits in detail. 

The Q of the coils of the discriminating network and the coupling of 
L s and Z/ 4 to L 2 must be such that the network will be able to pass a 
frequency band which is twice the maximum rate of flutter. There- 
fore the band width over which the network must be linear is from 
800 to 1200 cycles/sec. For linearity it is important that the mutual 
coupling between Z/ 2 and L 3 is the same as from L 2 to L 4 . For the same 
reason the product of L 2 C 2 must equal the product of (L 3 + Z/ 4 ) C 3 . 

Careful adjustment of the discriminating network cannot be stressed 
too much, as the successful operation of the flutter measuring set 
depends on it. 




The drift meter indicates rates between and 2-cycles and is made 
up of V6, RI, Rt, Pz, and M2. The meter is also used as a tuning 
indicator to adjust the oscillator frequency to produce a 1000-cycle 
beat with the incoming signal. PZ is used to adjust the plate currents 
of tube V6-1 and V6-2 to zero current through the meter M2 when no 
signal is applied. 

V7 is a cathode-follower to couple the high impedance output of the 
discriminator to the low-pass filter C 4 L 5 . 

Again the load resistance R$ is chosen of a value to make it equal 
to the output impedance of V7. The image impedance of the low-pass 
filter is the same as R& and the cut-off frequency is 250 cycles. The 
filter is a two-section constant-K type, and will attenuate the carrier 
frequency (1000 cycles/sec) by more than 50 db. 

+ 1.0 





Figure 4. 

PS and P 4 are two step potentiometers coupled by a common shaft, 
and are used to set the sensitivity of the flutter measuring set. They 
are set for 0.1, 0.3 and 1.0% flutter and drift measurements. 

Figure 4 shows the response curve of the discriminating network 
when driven from a 10,000-ohm generator of constant amplitude. 
From this curve it is seen that the output is sufficiently linear with 
frequency to obviate the necessity for compensating circuits. When 
tested dynamically with a frequency-modulated audio-frequency 
oscillator (described by P. V. Smith and Ed Stanko in this JOURNAL 
in March, 1949) it was found that no nonlinearity existed. 

The output from the low-pass filter is fed directly into the final 
amplifier. This amplifier has a frequency response uniform within 1 
db from 1.5 to 200 cycles. Again negative feedback is used to adjust 
the gain and the output impedance of the amplifier. The plate voltage 
of the last tube is adjusted to such a value that the amplifier will work 
as a clipper at about 2 db above full-scale deflection of the meter. 
This is a very necessary precaution as transients frequently occur in 



flutter measurements which may burn out the thermocouple of the 

The network following the final amplifier consists of a low-pass, 
high-pass and band-pass filter. Figure 5 shows the frequency char- 
acteristics of the three filters when inserted between a 2600-ohm gen- 
erator and a 2600-ohm load. 

The percent flutter meter is connected directly across the output of 
the network switch. The oscilloscope output is connected across the 
meter terminals. Short-circuiting these terminals will also short- 
circuit the meter. If the input impedance to the oscilloscope is 
normal (over 100,000 ohms), the meter readings will not be affected 
and both the meter and oscilloscope can be used simultaneously. 

J 5 

10 30 50 /OO 



Figure 5. 

The rapid recording oscillograph output is designed to work into a 
500-ohm circuit. If the oscillograph is plugged in, it will automati- 
cally lift the meter circuit. 


Two of the flutter measuring sets have been built and give satis- 
factory service. The circuit is stable and independent of line voltage 
fluctuation. Flutter readings of as low as 0.02% total, 2- to 200-cycle 
rate, are reproducible. 


The writer wishes to acknowledge the very excellent assistance and co-opera- 
tion of the Hollywood Engineering Dept. of RCA Victor Div., and in particular 
the help of Kurt Singer, who played an important part in the final calibration and 
testing of this design. 

A Reflex 35-Mm Magazine 
Motion Picture Camera 



SUMMARY: A new portable professional 35-mm motion picture camera has 
been recently introduced into this country from France, embodying the fol- 
lowing characteristics: reflex shutter, adjustable from 200 to 40 (viewing 
is through the taking lens at all times); instantaneous loading of 400- or 100- 
ft magazines; divergent three-lens turret permitting use of 24- to 500-mm 
lenses without interference; interchangeable 6- or 8-volt electric motor, hand 
or spring drive; double pull-down ratchet movement with unique system of 
pressure pads and spring tensions to insure steadiness. The exterior shape 
of the camera is designed to fit the body, thus insuring steady hand held 
operation. The flat base cf the camera is made to fit the rapid mounting 
dovetail of the tripod head. 

WE HAVE TOGETHER designed the Camerette (Patents Coutant- 
Mathot) which last year was introduced in the United States. 
The Camerette is manufactured in Paris, France, by the Etablisse- 
ments Cinematographiques Eclair, manufacturers of the Camere'clair 
400-Ft and of the Camereclair Studio 1000-Ft cameras. The Camer- 
ette's name for the European market is "Cameflex." 

In 1944, during the German occupation, we decided to put our 
ideas together to create a really modern portable camera. Drawing 
upon the twenty years of experience we had in the motion picture 
industry, we each had specific ideas about the conditions a portable 
camera would have to meet to answer the needs of the cameraman and 
the producer. 

Our basic ideas have been patented, called "Patents Coutant- 
Mathot," and these patents cover the main Camerette features, most 
of which are completely new. We will be happy if we have succeeded 
in making the work of the cameraman easier, and if we have helped to 
improve motion picture camera techniques. 

The principal characteristics of the camera are as follows: 

General Shape. The shape of the camera with the 400-ft magazine 
attached is such that it can easily be hand held by resting the maga- 
zine on the shoulder, holding the camera by the motor with the right 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 





Fig. 1. Diagram of the Camerette. 

Front section 

Front aperture plate 

Base or seat of the turret 

Turret lock 

Standard filter holder for lens 

Lens mounting 

Silvered mirror placed on the front 

of the shutter blade 
Shutter and its mounting 
Ground glass 
Flat camera base, fitting the special 

dovetail in rapid mounting tripod 


Magazine engaging bolt 
Groove for magazine lock 
Light traps which are automatically 

opened when magazine is attached 

to camera 

15. Sprocket 

16. Take-up hub with spring clips ac- 

cepting standard film spools, male 
or female 

17. 400-ft automatic film gate magazine 

18. Lens shade shown on 100-mm lens 

19. Rear channel of film on magazine 

20. Film passage to take-up spool 

21. Top pressure pad maintaining film 

No. 23 in proper position at the 
aperture No. 2 

22. Bottom pressure pad keeping the 

film properly aligned for the pull- 
down claws No. 24 

23. Film, upper loop 

24. Pull-down claws 

25. Tempered steel pad for the lateral 

spring guides 


hand, using the left hand to focus the lens, and support the camera. 
The balance and shape are such as to give extremely steady hand-held 
operation since the camera is held close to the body, and firmly sup- 
ported. The flat base of the camera (Fig. 1-11) and the eyepiece 
swinging into the vertical viewing position allow for placing the camera 
on the ground without the use of any support. Use on the tripod 
is equally rapidly accomplished by means of the rapid mounting dove- 
tail head. 

Viewing is through the taking lens by reflection from an unbreak- 
able front silvered mirror placed on the front of the shutter blade, and 
rotating with it (Fig. 1-7). This image is transmitted by a ground 
glass (Fig. 1-9) and prisms (Fig. 1-10) to the magnifying eyepiece. 
The eyepiece is fitted with a calibrated focusing adjustment, and can 
be set in three positions: horizontal for normal use, vertical when 
the camera is used from a very low angle, and, when not in use, the 
eyepiece is placed in the lowered position for storage in the carrying 
case. A self-closing light guard is provided for use in direct sunlight 
or other strong light ; under normal lighting conditions its use is not 
required. The camera can be obtained with either right or left eye- 
pieces. This reflex method of viewing eliminates the necessity of 
auxiliary finders, has the advantage of accurate framing with no 
parallax, and makes it possible to follow focus visually during actual 

Magazines. The automatic film gate magazines are available in 
either 400- or 100-ft film capacities, and can be used interchangeably. 
Magazines are instantly attached and locked to the camera unit by a 
simple pressure of the hand. Unlocking is equally rapid by pressing 
the locking knob and removing the magazine. Magazines can be 
changed while the camera is in operation. The camera has its best 
balance for hand-held operation with the 400-ft magazine, since the 
magazine can be supported on the operator's shoulder, permitting 
steadier work with less fatigue, because of the excellent weight dis- 

There are two fiber pressure pads on the front of the magazine, the 
upper pad maintaining the film in the proper position at the aperture, 
the lower one insuring steadiness, guiding the film in its relation to the 
pull-down claws, situated on the camera proper. The double pull- 
down claws, pressure pads and guides which keep the film traveling 
in a straight path past the aperture insure absolute steadiness. 

The preloaded automatic magazines offer a distinct advantage in 
the saving of production time due to waits for reloading. For cold 
weather operations where loading becomes a problem because of the 







necessity of wearing thick fur gloves, the preloaded automatic film 
gate magazine is the obvious solution. Loading the magazines is 
simple, and can be done in either darkroom or changing bag. The 
loops are formed in daylight, and are not critical. Film wound either 
emulsion in or out can be used. 

Figure 4 shows the method of attaching the magazines to the cam- 

Fig. 4. Attaching the film magazine. 

Movement and Aperture Plate. The movement consists of two ratchet 
pull-down claws engaging the film, which is kept in its proper position 
by the two pressure pads on the film gate magazine. The top pad is 
designed to keep the film flat and in the focal plane. The surface of 
the pressure pad and the system of pressure applied are such as to 
avoid all pressure against a hard surface, thus preventing distortion 
of the film at the aperture. The bottom pad, which holds the film 
only at the edges, keeps the film properly aligned for the pull-down 
claws, insuring absolute steadiness. The lateral guidance is assured 
by two spring guides in the magazine, placed in the curved parts of 


the film loops. By keeping this pressure in the loops, warping or 
twisting of the film is prevented, since there is more resistance to 
lateral pressure in the curved sections. The elimination of aperture 
pins permits simpler, more compact construction, with consequent 
freedom from mechanical failure and repairs. Another advantage of 
this type of movement is the camera's adaptability to extreme changes 
of temperature, since wide variations in film and perforation size can 
be tolerated. 

The aperture plate is made of one piece of stainless steel, hand 
polished and undercut to prevent scratching. With the magazine 
removed, the plate is readily accessible for checking and cleaning. 
A special guard is provided for the aperture plate to prevent damage 
when the camera is dismounted for packing. 

Shutter. The shutter blade, in front of which is placed the reflex 
mirror, has a maximum aperture of 200. This is adjustable to 
40 by means of a graduated shutter disc (sliding behind the reflect- 
ing mirror); its position is controlled by an exterior knob. The 
Camerette Model C has a 230 shutter, adjustable to 110. See Fig. 

Drive. There are three alternative drives for the camera: electric 
motor, spring motor or hand-gear box. The change from one to the 
other can be rapidly and easily accomplished. The electric motors 
and hand-gear boxes are placed on the side of the camera at the right 
hand of the operator. The spring motor attaches to a special support 
on the back of the camera and rests beside the magazine. The stand- 
ard electric motor serves as the handle for the camera. 

The starting and stopping switches are in front and are operated 
by the little finger of the right hand. Speed control is obtained 
by turning the rheostat knob on the top of the motor with the thumb 
of the right hand, while holding the release catch with the index 
finger. This motor operates on either 6 or 8 volts of direct current 
supplied by a set of lead batteries. The batteries are mounted in a 
leather waist belt for carrying, and weigh 9 Ib. An electric charger 
operating on either 110 or 220 volts is part of the standard equipment. 
The batteries have a capacity of 15 amp-hr and will operate ten 400- 
ft magazines, or 4000 ft of film, on one charge. The motor switch has 
three positions, the middle position cutting out the rheostat for a 
fraction of a second to help overcome starting inertia, enabling the 
camera to come to speed quickly. This position can be utilized also 
to change the camera from high to normal speeds without changing 
the setting of the rheostat. 


The spring motor, used only in emergency, is entirely ball bearing 
mounted, and will resist extremely cold temperatures. It is capable 
of running 45 ft of film on each wind. There is a button for speed 
regulation, and a crank for winding. 

The hand-gear box attaches in the same position as the electric 
motor. There are three gear ratios, one, eight, or sixteen frames per 
revolution of the crank. A 220- volt, 50-cycle synchronous motor is 
also available. 

Turret. The divergent three-lens turret is designed to accommodate 
lenses from 24 to 500 mm without cut-off. The standard mounts are 
of the bayonet type, fitted with grips for focusing, and lens hoods 
with spring clips in which either the metal lens caps or filter holders 
can be inserted. Mounts are available for any standard lenses. 
Normally the camera is supplied with Kinoptik lenses which are 
coated F2 apochromats available in focal lengths from 25 to 500 mm. 
(See Figs. 1-5 and 1-6.) 

Filters. Three types of filter mounts are supplied. 

1. Round filters which clip inside the sunshade of each lens (di- 
ameter of filters 40 to 75 mm). 

2. Gelatin filter holders placed behind the lens in the aperture. 

3. Regular Wratten 3-in. square filters can be used in the matte 
box, which will accommodate two such filters. (Fig. 1-5 shows 
round-type filter holder incorporated in lens house.) 

Tachometer. The camera is provided with a magnetic tachometer, 
graduated from 8 to 40 frames/sec. 

Tripod and Tripod Head. The special tripod supplied with the 
camera is made of dural, and weighs only 13 Ib with the normal legs. 
It measures, closed, 3 ft 5 in., and 5 ft 6 in. fully extended. Medium 
and baby legs are available and can be easily attached to the tripod 
head. The tripod head can be rapidly detached from the legs by 
means of a clamp. An auxiliary clamp is also provided, permitting 
fastening the camera on any type of support. The camera is mounted 
on the tripod by means of a dovetail which receives the flat base of the 
camera, and locked into place by a spring bolt. Pulling down on the 
spring bolt and starting the camera out of the dovetail by means of 
the cam lever on the tripod head frees the camera from the tripod. 
A pan handle is carried on a clip on the tripod. 

The exterior of the camera is protected by a black anodized finish, 
which is very durable, weather and shock resistant. The camera, with 
motor, three lenses and 400-ft magazine, weighs only 14 Ib. 

Economy in Small- Scale 
Motion Picture Lighting 



SUMMARY: There is an apparent need and trend toward reducing the 
amount of electric power required for illumination in motion picture produc- 
tion. Although the interest in this situation extends through the entire in- 
dustry, the greatest economic significance probably concerns the smaller 
nontheatrical producers, many of whom are working with direct 16-mm film. 
For present purposes, this report considers only the problem of small sets and 
field work. 

IN A GENERAL CONSIDERATION of the lighting equipment required 
by small producers, several factors are immediately evident: 
(1) dependent on the location and size of the set or scene, the lighting 
required may vary considerably in both type and volume; (2) the 
quantity of lighting units available among small producers varies to 
great extremes ; (3) the economy of operation is an almost universal 
problem ; (4) the art of scene lighting is affected by the lighting units 
available; and (5) there are many electrical factors which must be 
given due consideration. 

Specifically, such problems arise as how to accomplish filming with 
minimum power, how to decrease the expense of new power installa- 
tions, how to provide an increased over-all lighting for the slower 
color films without greater amperage, how to acquire quantity light- 
ing without adding many expensive lighting units, and how best to 
distribute the available electric power service from the viewpoint of 
economy. These questions are common to many small producers. 
Before attempting suggestions which may provide answers to some 
of these and other problems, it seems important to consider lighting 
in general. 

During the rapid development of the theatrical film industry, 
standard lighting equipment evolved which in its entirety provides 
dozens of types, styles and wattages of illuminants. Two principal 
classifications, arc lamps and incandescent lamps, can be subdivided 
into a host of forms among the variously sized flood lamps, spot 
lamps and intermediates. This complexity of lighting equipment is 
justifiable within the major film studios. It is not possible, however, 

A CONTRIBUTION: Submitted April 16, 1950. 



for the small 16-mm producer to emulate such conditions, nor is 
that necessarily desirable. It is then a question of selecting the least 
number of lighting units which can provide adequate illumination 
for the scope of production involved. In simplifying the range of 
lamps, a primary consideration is that of achieving the possibility of 
well-balanced and well-modeled scene lighting. In the same sense 
that the creation of good lighting is an art, so are the lights the tools 
of the trade. In most instances it is critical that (1) a means be 
provided for creating a general level of all-over lighting, (2) that in- 
dividual lights be available for specific modeling and accents, and 
(3) that a means is provided for creating high levels of directional 
penetrating light to create sunlight effects, produce specific shadows, 
and so forth. 

Figure 1 contains a group of sketches illustrating a number of im- 
portant basic types of lights. A present conclusion is that almost 
any producer will require certain of these units and particularly a 
number of adjustable spotlights and versatile flood units for model- 
ing purposes. 

There are many combinations of lighting equipment encountered 
among small producers. Filming has been done with a few simple 
reflectors on collapsible stands. More commonly, a small studio will 
have an assortment of fresnel lens keg spotlights, a number of open 
floods in various reflectors, and one or more high-wattage arc or in- 
candescent spot units. Unless numerous standard floods are grouped 
to provide a basic level of illumination, the greatest problem usually 
appears as the need for a form of over-all gross lighting which is dif- 
fuse, of high intensity, of the lowest possible amperage and the least 
in expense. 

One such system which seems to be acquiring popularity is low- 
amperage Colortran. 1 Another method, as devised by the author, 
employs banks of reflectorfloods in much the same manner as used in 
certain parts of television lighting. 2 


It is well known that the "photoflood" type of lamp, when burned 
at 110 to 120 v, will produce far more effective wattage per ampere 
than conventional incandescent lamps. The issuance of the re- 
flectorflood lamps, which provide a built-in directional distribution of 
all their light, has proved to be of considerable value in the problem 
of concentrating large amounts of illumination in restricted areas. 
With a built-in reflector, the need for a larger controlling reflector is 
eliminated and an important gain is made in both space and expense. 




Fig. 1. Rough sketches of some basic types of lights: A, scoop; B, sun arc; C, 
arc spotlamp; D, rifle lamp; E, strip light; F, small spot with focusing snout; 
G, incandescent sunspot; H, open bowl flood; I, broadside; J, senior solarspot; 
and K, baby keg. 


Fig. 2. "Exploded" view of lamp bank layout. 




By grouping twelve such reflectorfloods into a bank, a highly potent 
source of light providing the equivalent of 18,000 w may be secured 
from an area of about 2 X 3 ft. Each such bank draws only 60 amp 
as contrasted to approximately 180 amp of normal lighting. The 
weight of cable required is drastically reduced both from the power 
source and to the individual lamps. Another added advantage is the 
simple and inexpensive means of constructing the lamp bank. 

In the case of most good quality standard lighting equipment, there 
is no cheap substitute. With reflectorflood banks, however, it is 
very easy to provide an inexpensive assembly which lacks nothing in 

Fig. 3. Complete lamp on Saltzman stand. 

Fig. 4. A, simple series-parallel 
circuit with double-pole-double- 
throw, center-off switch; B, six 
lamps in parallel against six others. 

utility. Lightweight cast aluminum condulets with appropriately 
placed and threaded connection points (Fig. 2) can be spaced by long 
7- or 8-in. end-threaded nipples. To accommodate final assembly, 
strap bars can be added to the two central vertical rows. A framing 
yoke of conduit tubing can allow for angular positioning. The 
central yoke attachment to a stand can be constructed to permit 
rotation of the entire bank. This method of assembly provides all 
the necessary movement and setscrews may be used to retain locked 
positions. The stand itself may be of any desired type. In the 




present instance, mobility was critical and the very light and rugged 
Saltzman collapsible magnesium stands were secured (Fig. 3). These 
stands have a remarkably long maximum extension and a firm base. 
Thus for a total material cost of about $60, which includes stands, 
lamp banks of 18,000 w can be secured. Five such banks used for dif- 
fuse fill-type of illumination can provide 90,000 w for only 300 amp. 
If this same area were lighted by the use of individual or strip 2,000- 
lamps, 45 bulbs drawing about 800 amp and occupying considerable 
space would be required. 

Fig. 5. Plan for a master plugging box based on series-parallel circuit and em- 
ploying master switch. Switch "A" is Westinghouse Class 11-210-MS2 relay type 
operated by remote high-low-off push-button station. Switches "B" are Westing- 
house reversing drum type N-103-E. Switch "A" may also be obtained for 220 v 
where input current is three-wire 220-1 10-v. 


One of the principal objections to the use of the photoflood-type 
light is its relatively short life. The expense as well as the nuisance 
of constant replacement has limited its use in professional filming. 
The answer to these problems is found in the use of dimming units 
whereby the lamps may be burned at reduced brightness during 
preparation and between takes. Since adjustable rheostats and 
transformers would be heavy and costly, the simple method of high- 
low switching by a series-parallel circuit appears as the most econom- 
ical system. 




Fig. 6. Above, portable 300-amp capacity plugging box; 
below, master relay switch. 

In the use of a series-parallel circuit, a double-pole-double-throw, 
center-off switch may be employed as in Fig. 4a. Since it would be 
impractical to use a switch for each pair of reflectorflood lamps, a 
bank of twelve lamps may be divided so that six are balanced against 
six while in series (Fig. 4b). To eliminate the confusion of multiple 
switching points, each pair of cables may be brought to a centrally 
located portable switching box. It is difficult to secure snap switches 

186 ARTHUR L. SMITH August 

of the double-pole-double-throw type which will carry a total load 
of 60 amp and the use of knife switches is not convenient. A drum 
switch, such as the Westinghouse Type N-103-E, can be easily 
modified to serve, is quite inexpensive, extremely rugged and of ade- 
quate rating. 

The use of series-parallel switching may also be applied to other 
units employed on the set. Their cables also may be brought to a 
central plugging box and all the lighting controlled from a single area. 

Where labor, time and efficiency are critical, a further refinement 
may be added. By using a special master relay switch operated by 
remote control, it is possible to have all of the lighting units in use 
controlled by a single high-low-off push-button station. Although 
the use of simple series-parallel circuits is quite common, the addi- 
tion of a master switch seems to have been largely ignored in motion 
picture lighting. Several years ago, the author devised a circuit of 
this type 3 which has since been revised to permit amperages up to 
300 amp per master switch (Fig. 5). If more than 300 amp are re- 
quired, extra master switches may be added and controlled by the 
same push-button station as the original. 

In professional use, the plugging box and master switch (Fig. 6) 
are found to have notable advantages for saving time. Where small 
crews are employed, the entire set lighting may be culminated in a 
single box and switched to bright, dim or off with the touch of the 
cameraman's or director's finger. The cost of the switches for a 
master control plugging box accommodating sixteen lamps (for 
example, three twelve-lamp banks and ten other units) by remote 
control would be less than two hundred dollars. If the drum switches 
for individual pairs are eliminated, the cost will be much less than one 
hundred dollars. 


A problem which is related in economy and efficiency to lighting, 
concerns the electrical materials and factors involved in providing 
power. Although the time-honored Kleigl and other types of stand- 
ard connectors and branch-offs seem still to be in predominant use, 
there seems also to be some justification for considering other devices. 
Any production set electrician who has used nonlockable plugs has 
doubtlessly experienced unplanned disconnections. Inexpensive 
Twistlock connectors will carry up to 20 amp and easily overload to 
30 amp. Their use is so simple and their positive locking action so 
durable that it is surprising not to encounter them more often. When 
loads of greater power are to be carried, Hubbelock connectors may 




be employed. Some of the four-contact type with pairs barred to- 
gether are rated at 70 amp and have carried over 150 amp without 
trouble for the author. These connectors are available in styles 
which are watertight and for certain field work this advantage is 
critical. Four-prong Cannon-type plugs rated at 200 amp have 
been adapted satisfactorily. 

Many producers decrease cable size and cost by bringing main 
power lines to a distribution point on a three-wire 220-1 10-v circuit. 

Fig. 7. Lamp banks assembled and ready for use in small studio scene. 

This device may also be used in connection with a remote control 
plugging box, half of the current going to each side of the box. In 
this instance, the weight of the cable will, of course, bear a relation 
to the capacity of the box and although an open temporary cable 
may be somewhat overloaded, an extreme in this regard may be 
dangerous as well as decrease the input voltage and cause a lowering 
in the intensity and color of the lamps. Fortunately voltage drops 
in main line cables are not as great with 220 v as with direct 110 v and 
this is an added advantage of the three-wire 220-1 lO-v system. 


If power is to be supplied through direct lines without series- 
parallel switching, a very durable, safe and inexpensive service can be 
obtained with the use of a breaker panel. For example, a panel hav- 
ing twelve 50-amp breakers can be used with twelve Hubbelock re- 
ceptacles mounted in the gutters. By a curious coincidence an en- 
tire panel may be secured with optional breaker capacities up to 50 
amp each without adding to the basic price of about sixty-five dollars. 
The main bus bars can also be specified of a weight to permit distri- 
bution of a total load of around 400 amp. 

Many of the various electrical connectors and power boxes re- 
quired for heavy amperage are extremely expensive. It is for this 
reason that it seems important to note those particular materials 
which are rugged, suitable and least expensive. 


An attempt has been made to outline various means by which a 
small producer may secure a reasonably professional quantity and 
quality of motion picture illumination on a basis of added efficiency 
and economy. Standard types of lighting have been retained for 
modeling, accent, high key and character. The basic level of il- 
lumination is obtained by reflectorflood banks which in turn are 
made efficient by means of a master switching box. This system of 
illumination provides added wattage by inexpensive means and the 
critical art of lighting employs essentially the same "tools" as those 
of the standard large scale studio. Heat from set lighting may be 
reduced by dim settings of either pair of switches or the master switch 
except for the time of actual shooting, thus providing comfort for the 
cast and extending lamp life by five or ten times. 

A great many of the new and smaller production services are faced 
with the high acquisition cost of good-quality lighting units on today's 
market. Rather than sacrifice picture quality or operate largely 
with open lens apertures, it seems desirable to utilize a type of light- 
ing which possesses both economic and professional advantages. 


1. "Packaged illumination," Cinematographer, Feb. 1949. 

2. W. C. Eddy, "Television studio lighting," Jour. SMPE, vol. 49, pp. 334-341, 

Oct. 1947. 

3. Arthur L. Smith, "A light dimming unit," J. Riol. Phot. Assn., vol. 9 pp. 206- 

209, June 1941. 

Component Arrangement for a 
Versatile Television Receiver 





SUMMARY: The requirements and the arrangement of the various elements 
of a rather versatile television receiving system are described. A commercial 
unit is used to illustrate some of the practical applications of ideas suggested 
in this paper. 

/COINCIDENT WITH large scale utilization of home television re- 
^^ ceivers, there has been an increasing demand for a television 
receiving system of a somewhat more elaborate nature, for a larger 
and more critical audience. 

The demand of such a system can well be understood when one 
stops to consider the potentialities of television programs, not only 
for entertainment purposes, but also for educational and industrial 
uses. Several examples of possible applications of such a system can 
best be represented by the present-day use of both the Army and 
Navy of television for training and experimental programs. Philadel- 
phia and other school systems are using television programs as an 
additional aid in their educational systems, and recently the medical 
profession has applied television for viewing surgical operations. 

The utilization of such a system will be, in general, by an audience 
larger than can be accommodated by a single home-type receiver, and 
thus it is inferred that a projection type unit will be employed. The 
purpose of this paper is to discuss the arrangement and considerations 
given to the elements of such a system and later on to describe a unit 
employing some of these basic concepts. 

Figure 1 shows a block diagram of a possible receiving system. 
This diagram describes a somewhat elaborate installation, showing a 
few of the possible combinations that can be achieved with this ar- 
rangement. However, by proper choice of basic elements this system 
can be reduced in magnitude and still provide suitable service, de- 
pendent on the size and requirement of the installation. 

It should be noted from Fig. 1 that the system is broken down into 
groups of individual chassis, each performing a basic function of a tele- 

PRESENTED: April 25, 1950, at the SMPTE Convention in Chicago. 





vision receiving system. Certain ones of the chassis can be removed, 
modified, or added to the system in order to meet the requirements of 
the individual installation. Each of the principal chassis could obtain 
its own power supply so that its removal or addition would not affect 
the operation of the over-all system. This arrangement also simplifies 
testing and servicing. 

The antenna lead-in cable is shown connected to the control box 
as the r-f (radio-frequency) tuner is situated in this unit. Locating 
the r-f tuner in this manner makes it possible to provide remote 
station selection without resorting to an elaborate servo or step-type 
mechanism. This feature is accomplished by connecting the out- 

Fig. 1. A receiving system. 

put of the mixer stage to the first i-f (intermediate-frequency) stage 
by means of a low-impedance link, coupled through shielded coaxial 
cable. If the unit is to be operated on other than the standard 
channels, it is only necessary to exchange the r-f tuner for one match- 
ing the desired frequencies, the only proviso being that the inter- 
mediate frequency of the system remain the same. 

The r-f tuner should be of such design that a booster amplifier 
should not normally be required. Maximum useful sensitivity con- 
sistent with a good signal-to-noise ratio and unwanted signal rejection, 
without a sacrifice to bandwidth, are the prime requisites of the r-f 

The remaining controls (also located in the control box), such as 


contrast, brightness, tone, volume and the master power switch, should 
operate on d-c voltages wherever possible in order to reduce the 
amount of cable required at installation and also to eliminate the 
necessity of returning signal voltage to the control box, to avoid 
possible signal distortion. 

The composite sound, synchronization and video signal is naturally 
brought into the receiver and distribution chassis at intermediate 
frequency. The i-f stages of the receiver follow the usual design 
pattern, with special precautions taken to insure proper bandwidth, 
sufficient gain and ample trapping for adjacent channel interference. 

Based on past experiences, and with minor circuit changes, a con- 
venient input jack should be provided, to permit the input from a 
closed loop camera chain to be easily switched into the circuit after the 
second detector. 

In the system of this nature it is desirable to employ a well de- 
signed a-g-c (automatic-gain-control) circuit which would be located 
in the receiver and distribution chassis. Automatic gain control, 
besides reducing the effects of varying input signal levels, also helps 
to suppress random interference from noise signals. It is also of 
value in insuring that synchronization signals actuating the sweep 
circuits operate from the same relative level with each operating 
pulse. This type of operation serves to remove a certain amount of 
jitter in the picture, which would otherwise be noticeable. 

In order to provide for maximum flexibility in the choice and loca- 
tion of the sound and projection system, the sound, synchronization 
and video signals, after proper separation and detection, are applied 
to individual low-impedance line-driven amplifiers. Several of these 
low-impedance amplifiers can be operated in parallel so as to provide 
additional outlets to monitor or "slave" units. 

The size and method of producing the projected picture determine 
the ultimate design of the sweep chassis; however, certain funda- 
mental features should be incorporated in its specifications. Maxi- 
mum consideration should be given to the design of the sweep genera- 
tion circuits to insure linearity of sweep, stable synchronization and 
elimination of pairing, because minor irregularities, not noticeable on 
small-size screens, are quite evident when the picture is enlarged. 

To meet these requirements, it is advisable to utilize automatic 
synchronizing circuits in both the vertical and horizontal sweep gen- 
erators. In order to retain flexibility and also to remove a possible 
source of interference to synchronization, the high-voltage power 
supply should be divorced from the sweep chassis. 

The high-voltage power supply will depend on the type of projec- 


tion system employed. It should be well regulated with respect to 
variations in line voltage and average picture brightness. In order to 
reduce the danger of shock hazard to operating personnel and also to 
reduce the size of the unit, it is desirable to operate the high-voltage 
power supply at a frequency higher than line frequency. The energy 
storage capacity of the filter units required at power-line frequencies 
makes these units quite lethal. If an r-f type of high voltage power 
supply is employed, adequate shielding should be provided in order to 
prevent interfering signals being picked up by the receiver chassis. 

The video amplifier is placed in a separate chassis so that its size 
and location can be chosen to meet the type of projection unit em- 
ployed. If a separate amplifier were riot used, the distance between 
the projection equipment and the receiving equipment could be the 
limiting factor on the high-frequency response due to the shunt ca- 
pacitance of the projection tube. 

In order that the high fidelity transmission of the f-m (frequency- 
modulation) sound signal be used to advantage, the audio system 
should be of high quality to be capable of reproducing the audio- 
modulation range of 50 to 15,000 cycles/sec, which is used in the f-m 
transmission for television sound. Again, the power handling capa- 
bility and location of the unit will be dependent on the type of in- 

The viewing system is dependent on the picture-size requirement of 
the installation. Two methods are currently being used; one is by 
direct view from the cathode-ray tube, the other by projection from 
the face of the cathode-ray tube; however, other methods are possible, 
such as film recording and projection. 

The direct-view method is somewhat limited as to the number of 
people that can be accommodated by it. However, in some cases, 
this drawback can be circumvented by using a number of direct-view 
"slave" units. The technique used would be the same as that em- 
ployed in home television receivers. 

Figure 1 indicates a rear-view projection system. However, it is 
possible to use front projection with the optical barrel suspended from 
the ceiling, or mounted on a fixed or movable dolly. Projection from 
the face of the cathode-ray tube affords an excellent method for in- 
creasing the screen size; however, certain technical limitations place a 
limit on the size and brightness of the picture that can be obtained 
with this method. In order to obtain maximum results with this 
method, a highly efficient optical system must be employed. At 
present the Schmidt system, or variations of it, give the best results. 
Corrections should be made for distortion and aberration. 





The Precision Television Receiver demonstrates some of the prac- 
tical aspects of this paper. Although built as a packaged unit to 
simplify installation, it nevertheless incorporates some of the features 

Fig. 2. The Precision Television Receiver. 

discussed in this paper. Figure 2 shows the front view of the unit, 
and Fig. 3 shows the arrangement and interconnections of the various 
chassis used in this receiver. With minor changes, the components of 
this system can be extended to include the elaborate arrangement 
earlier described. 


The precision Television Receiver Model L-10 features a 27 X 36 
in. rear-projected picture. In order to obtain a picture of this size, a 
5TP4 projection tube with 30 kv applied to the second anode is used 
in conjunction with a Schmidt optical system. The optical system 
consists of an optical barrel supporting the projection tube, the 12-in. 
front-surfaced spherical mirror and the corrector lens. A 45-deg, 
front-surfaced mirror located in the top section of the cabinet directs 
the reflected light onto the rear of a plastic translucent screen having 
directional properties. The high-light brightness as measured on a 
standard test pattern has been measured at 30 ft-L and the resolving 
power of the optical system has also been measured to be somewhat 
better than 1000 lines. 

The r-f tuner is shown mounted on the receiver chassis but it can be 
conveniently removed to serve a remote location without changing its 
operating characteristics, because the output transformer used is de- 
signed so as to connect with the first i-f amplifier stage through a low- 
impedance link coupling. The cable used to effect the coupling can be 
any commercial, 90-ohm, shielded coaxial cable up to 50 ft in length. 

Channel selection is made by switching to individually tuned cir- 
cuits for each section of the tuner circuit. These tuned circuits are 
mounted on low-loss bakelite clips which snap into their individual 
sections on a turret selector. The sequence of selection can be easily 
arranged in any desired fashion. If desired, tuned circuits for operat- 
ing on channels other than standard can easily be inserted. This 
tuner is capable of 100-juv sensitivity at normal bandwidth with an 
average signal-to-noise ratio of eleven to one. 

The first two stages of i-f amplification are common to both the 
sound and picture signals. The sound signal is separated from the 
picture signal in the plate circuit of the second stage and after ad- 
ditional amplification and limiting, the signal is then detected in a 
frequency discriminator stage. The picture signal passes through two 
additional i-f stages before being detected. Six trap circuits are 
utilized in the picture stages to remove the undesired adjacent chan- 
nel and sound signals. After detection the video signal is applied to a 
cathode follower stage designed to match a coaxial cable feeding the 
remote video amplifier. 

A separate detection stage is utilized to separate a portion of the 
video signal for operating the automatic gain control and synchroniza- 
tion separation circuits. The synchronization signal is further ampli- 
fied before being applied to a cathode follower stage for line matching 
to the sweep chassis. The signal for automatic gain control is d-c 
amplified and filtered in three stages and is designed so as to give 




separate delay characteristics to the r-f and the i-f stages controlled by 
automatic gain control. The amplified a-g-c circuit in this receiver 
will equalize the second detector output for a range of signals between 
500 juv minimum and 40 mv maximum. 

The sound amplifier of this system is also shown mounted on the 
same chassis as the receiver; however, this unit can easily be divorced 
from the receiver chassis for remote-location purposes. The amplifier 
is capable of 20-w power output and when the bass and treble controls 
are cut out of the circuit, the response is essentially flat to 20,000 
cycles. At 20-w output the harmonic content at center frequency 

Fig. 3. Arrangement and interconnections of various chassis. 

C, optical barrel E, 30-kv high-voltage supply 

D, video amplifier F, speaker 

A, receiver and sound chassis 

B, sweep chassis 

has been measured to be 1J^%. For the same power output, the 
harmonic content for both the lower and upper part of the frequency 
spectrum has been measured at 2.2%. A 12-in. permanent magnet 
speaker, with a power rating of 25 w, is used to load the sound ampli- 

The sweep unit is shown as an additional chassis and contains its 
own power supply. It is only necessary to change the deflection out- 
put transformers of this unit in order to operate projection tubes hav- 
ing a higher second anode voltage than the 30 kv rated as maximum 
for the 5TP4 tube. The horizontal and vertical sweep oscillators are 


so well stabilized that under the worst condition of noise, where a 
picture is just barely visible through the snow, no trouble is en- 
countered in maintaining synchronization. 

The 30-kv high-voltage power supply is shown in its separate 
shielded chassis which includes its own low-voltage power supply. 
The unit operates on the r-f principle and develops the 30 kv by means 
of a voltage tripler circuit. The energy storage capacity is repre- 
sented in effective shunt capacitance of 250 /zjum. This small capaci- 
tance in series with a one-megohm resistor materially reduces the 
danger of shock hazards. The unit has a current capacity larger than 
is necessary for the operation of the 5TP4 projection tube. A 5-kv tap 
is taken off from a potentiometer and is used for electrostatic focusing 
of the projection tube. 

The remote video amplifier completes the system and makes it 
possible to operate the projection tube at distances up to 50 ft away 
from the receiver chassis without any detrimental effects upon the 
video signal due to the shunt capacitance of the projection tube. The 
gain supplied in this unit is sufficient to drive tubes having a higher 
second anode voltage than the 5TP4. 

Today, receivers of this type are being used as a part of the Army 
Reserve Training Program. Similar models have also been used with 
marked success in conjunction with school programs, and modified 
versions are being submitted to the Navy for approval for use in 
some of its training and experimental programs. 

Associated with the present demands for a more elaborate television 
system, it is felt that units in the near future will contain many of the 
considerations given in this paper. 

The authors gratefully acknowledge the collaboration of the follow- 
ing of the staff of General Precision Laboratory: Dr. R. L. Garman, 
Director of Research; T. P. Dewhirst, Project Engineer; R. Ander- 
son; and E. H. Lombardi, whose engineering reports represented the 
basic source material for this paper. 

Designing Engine- Generator Equip 
ment for Motion Picture Locations 



SUMMARY: Artificial lighting on outdoor motion picture sets is essential for 
both day and night photography. In most cases an electrical distribution 
system is not available at the selected location, and power must be supplied 
by electric generators driven by internal-combustion engines. Because 
standard, commercially available, engine-generator sets are not suitable for 
the special performance requirements encountered in motion picture pho- 
tography it is necessary to design and construct special equipment having the 
required features. This paper describes and evaluates the engineering 
factors involved, and illustrates how each of the desired characteristics was 
attained in equipment recently constructed. 


THE ENGINEERING FACTORS which must be considered in the de- 
sign of engine-generator equipment for supplying electric power 
for lighting on motion picture locations are as follows : 

1. Electric Power. The generated electric power should be 120-v, 
d-c, to supply satisfactorily both arc lamps, which require direct cur- 
rent, and incandescent lamps, which operate on either alternating or 
direct current. Experience has shown that engine-generator sets 
having a capacity of between 750 and 1,400 amp 1 will currently 
satisfy the load requirement in practically all cases, with those at or 
near the higher rating being more in demand. The increase in the 
number of color pictures being made on locations indicates a possible 
future demand for engine-generator sets having capacities above 
1,400 amp, and a few sets capable of producing 2,000 to 2,500 amp 
have been constructed. However, in considering the feasibility of 
these larger units the advantage of increased capacity must be 
weighed against the disadvantage of decreased portability due to 
added size and weight. Also the relatively fewer occasions on which 
they can be used should be considered. To date, it has been generally 
accepted that it is more feasible to use two or more smaller, lighter, 
and hence more maneuverable, engine-generator sets to supply the 
higher current demands on location. 

A three-wire d-c system is superior to a two-wire system for distrib- 

PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. 


198 HANKINS AND MoLfc August 

uting large amounts of power. 2 A three-wire system produced by a 
single three-wire generator is not well suited for motion picture work 
because of the inherent variation in voltage caused by load un- 
balance, and such a system is best obtained by two generators con- 
nected in series. Driving two generators from one engine presents 
mechanical problems. When deciding upon the system to be used, 
the electrical advantages of the three-wire system should be weighed 
against the necessary mechanical complications required to produce 
it, and the decision is usually determined by the power rating. For 
engine-generator sets of capacities up to 1,400 amp which are to be 
driven by a single engine, experience has shown that it is better to 
use one two-wire generator and accept the disadvantages of the re- 
sulting two-wire distribution system. However, when considering 
larger loads up to 2,500 amp, the electrical advantage of the three- 
wire system becomes the governing factor, and a construction having 
two 120-v generators in series is to be preferred. Such plants should 
have two engines, one for each generator, thereby gaining an advan- 
tage in the form of flexibility, in that only one engine need be operated 
when the plant is supplying half -load or less. 

Good, consistent lighting is dependent upon close control of the 
voltage at the load-end of feeder cables. The lighting load en- 
countered on locations varies from a small fraction of the generator 
rating up to its maximum capacity. Feeder cables may be short or 
relatively long, dependent upon the conditions at the location. An 
appreciable increase in voltage cannot be tolerated because of the 
danger of damaging incandescent lamp filaments. Automatic volt- 
age regulation to meet these operating conditions is essential. 

Commutator ripple in the d-c voltage causes noise emission from 
carbon arcs which can be objectionable on sound sets. Generators 
for motion picture work should be so designed that their ripple volt- 
age does not exceed =*= J^ of 1% of rated voltage. Even then the 
ripple is, on many occasions, further reduced by means of filter cir- 
cuits using choke coils and capacitors. 3 

Since the duty cycle of an engine-generator set is of an intermittent 
nature, a generator which will produce the required maximum amount 
of power for approximately one-half hour without injurious heating is 
considered to be adequate. Hence, much smaller and lighter weight 
generators can be used than would be the case if it were necessary to 
give them a continuous rating at the maximum output. 

2. Prime Mover. Either a gasoline or diesel engine can be used to 
drive the generator and both types have been successfully em- 
ployed. The speed-power curve of the engine should match the 


generator requirements, bearing in mind that, as a protective meas- 
ure, engine horsepower should be somewhat below that capable of 
driving the generator at an injurious load. 

The engine should be equipped with a governor capable of manual 
adjustment to maintain automatically any desired speed within the 
generator operating range. 

3. Control. The controls for both the engine and generator should 
be conveniently grouped so a single operator can quickly perform all 
operating functions required. 

4- Noise. On sound locations the engine-generator set must be 
positioned so that its operation noise does not interfere with produc- 
tion. A design which effectively reduces the noise level saves setup 
time and permits the use of shorter feeder cables since the plant can 
be located closer to the action. 

5. Portability. The engine-generator set should be as small and 
light in weight as possible. Its dimensions should allow passage 
through door-openings in railway cars, and be suitable for mounting 
on a truck or trailer. Maneuverability in and out of " tight spots" 
on locations is essential. 

6. Dependability. More often than not, engine-generator sets 
operate in remote locations where supply parts and the repair-shop 
type of maintenance service are not available. A breakdown on such 
locations would hold up an entire company and increase production 
costs. For that reason the construction should be as foolproof as 
possible, using the minimum number of parts to satisfy operational 
requirements. The components employed should be of a standard 
commercial type which have proven their reliability under service 

7. Protective Devices. Automatic-operating safety devices should 
be incorporated to protect the engine-generator set against damage 
caused by abnormal conditions. 

8. Maintenance. The use of standard, readily available, conir 
mercially proven components greatly reduces the maintenance 
problem. In addition, the engine-generator set should be designed 
so that those parts which require periodic maintenance are readily 


In order to illustrate methods which may be employed to meet the 
foregoing basic requirements, a particular design of a 150-kw engine- 
generator set recently manufactured by the Mole-Richardson Co. is 
described. The completed unit (Figs. 1 and 2) consists of a cubicle 








Fig. 1. 150 Kw engine-generator set. 

54 in. wide by 72 in. high by 118 in. long and weighs approximately 
10,000 Ib. The enclosure forms the outside walls of three separate 
internal compartments: the generator compartment at the radiator 
end of the plant, the engine compartment at the rear of the plant and 
the muffler compartment at the top. All outer surfaces of the hous- 
ing are polished, stainless steel. The design features directly related 
to the basic requirements are described as follows : 

1. Electric Power. When contemplating the construction of a 
power-package, the electrical requirements and the engine must be 
simultaneously considered. Obviously, a special engine cannot be 
designed and built for this particular application, and one which will 
drive a generator of the approximate desired rating must be selected 
from those commercially available. Generator performance should 











Fig. 2. Oblique view, off-side and rear. 

conform to the speed-horsepower characteristics of the chosen engine, 
and modification of standard generator design is generally required. 
In this instance, after consideration of available engines, it was 
decided to construct equipment capable of delivering 1,200 amp at 
125 v, or 150 kw. To meet this requirement, a generator (Fig. 3) 
was chosen which has an intermittent duty rating of 1,400 amp, 125 
v at 1,800 rpm, and a continuous rating of 1,000 amp, 125 v at 1,400 
rpm. It will pick up its voltage at 1,350 rpm without exceeding rated 
field current. The voltage rating allows for a 5-v line drop when 
maintaining 120 v at the load-end of feeder cables. The generator is 
two-wire, self-excited, flat compounded, Class B insulated, and weighs 
approximately 2,500 Ib. Its voltage ripple characteristic is within 
the allowable limit of == J^ of 1% of rated voltage. An automatic 








Fig. 3. Engine-generator assembly. 


voltage regulator, located as shown in Fig. 2, maintains the proper 
voltage setting over the speed range of the generator. A field con- 
trol rheostat is provided so that voltage can be manually controlled 
in case trouble should develop with automatic voltage regulation. 
A field control switch at the control panel permits the operator to 
select the type of voltage regulation desired. 

2. Prime Mover. The engine (Fig. 4) selected for this application 
develops 275 hp at 1,800 rpm as shown by its speed-horsepower 
curve (Fig. 5). After allowance is made for the power consumed by 
auxiliary drives, the engine is nearly fully loaded by 150-kw generator 
output at 1,800 rpm. Likewise, the engine power-input conforms to 
generator power-output at 1,400 rpm. The load current can there- 
fore be increased from 1,000 amp at 1,400 rpm to 1,200 amp at 1,800 
rpm at a rate of about 50 amp for each additional 100 rpm. This is a 
desirable feature since there is no need to operate the engine at a 
speed higher than necessary to develop the required horsepower. 




Fig. 4. Engine-generator assembly. 

The engine is of the industrial type having six cylinders with a 
bore of 5% in., a 7-in. stroke, a displacement of 1,090 cu in., a com- 
pression ratio of 5.7 to 1, and a weight of approximately 2,600 Ib. It 
is designed to consume ethyl gasoline fuel having an octane rating 
of approximately 80. 

This engine is equipped at the factory with a governor adjusted to 
maintain an engine-operating speed of 1,400 rpm. The tension of an 
added external governor spring is varied by a knob at the control 
panel so engine speed can be adjusted above that which would other- 
wise be maintained by the governor proper. This allows the oper- 
ator manually to adjust the automatic governor over the rated speed 
range of 1,400 to 1,800 rpm. 

The generator end bell is designed to flange mount with a rabbet fit 
to the flywheel bell housing of the engine. The special resilient-type 
coupling (Fig. 6) is designed for connecting the generator shaft to 
the engine flywheel. The flywheel is modified to accommodate its 
portion of the coupling. This type of construction results in a mini- 




mum of runout between the axis of rotation of the engine crankshaft 
and that of the generator armature and, therefore, insures long 
trouble-free coupling life. Mounting the generator direct to the 
engine bell housing results in a minimum over-all length of the coupled 

The engine and generator coupled to form one integral unit are 
mounted to the main-base frame primarily by means of the industrial 





^ i - 

























- r\ i 1 

: is 
> T 


eT i 

\ s 

V 1 



D f\\ 


r4 L 

3 ER 

A D < 


r- r> 









I ( 

H 1 

) o 


) Fl 

r\ Ul 


l r 




















1800 2200 

1000 1400 
R P M 

Fig. 5. Speed-horsepower curve for Hall-Scott model 400-0 engine. 

base of the engine (Figs. 3 and 4). Thick rubber belting under the 
engine base and rubber bushings and washers at the holddown bolts 
prevent metal-to-metal contact between the engine and the main-base 
frame, minimizing the transmission of engine vibrations to the power 
plant structural members. Since the generator is an overhung weight 
mounted to the engine bell housing, secondary supports consisting of 
rubber tube-form mountings (Fig. 3), are positioned between the 
main-base frame and the ears of the generator with their upward 









Fig. 6. Coupling. 

thrust forces evenly adjusted for minimum strain on the engine bell- 
housing. Here again, there is no metal-to-metal contact with the 
main-base frame. 

3. Control. All operating controls and adjustments of the engine- 
generator set are located on the control panel shown in Fig. 1 and il- 
lustrated in detail in Fig. 7. The controls and instruments are 
divided into two groups: the engine controls on the left and the 
electrical controls on the right. 

4. Noise. The major portion of operational noise of an engine- 
generator set consists of engine mechanical noise, exhaust noise, 
radiator fan and cooling air noise, and carburetor-intake noise. 

As described above, the engine and generator are resiliency 
mounted so there is no metal-to-metal contact with the structural 
members of the unit, thus minimizing the transmission of engine vi- 




















Fig. 7. Control panel. 

brations to the housing. The engine is completely enclosed by the 
engine compartment, the walls of which are insulated against sound 
transmission. The wall construction consists of the following, 
enumerated in order of position from outside to inside : stainless steel 
sheet, sound-deadening undercoating, air space, sound-deadening 
undercoating, asphalt compound, fiber glass insulation, fiber glass 
cloth and perforated stainless steel sheet. (Since the engine is com- 
pletely enclosed, cooling is dependent entirely upon the water-cooling 

The exhaust noise is reduced by a large muffler 14 in. in diameter 
by 6 ft long, mounted in the muffler compartment above the engine 
compartment. The walls of the muffler compartment are also in- 
sulated against sound transmission in the manner described above. 





Fig. 8. View through off-side generator 
compartment service door. 

A right-angle exhaust stack (Figs. 1 and 2) directs the exhaust gases 
in an upward direction. 

The attainment of adequate engine cooling with a minimum noise 
from fan and air flow is primarily accomplished with the use of a 
radiator having an exceptionally large frontal area (Figs. 1 and 8). 
The radiator has approximately twice the capacity of those normally 
used in industrial applications of this engine thereby providing suf- 
ficient cooling with a relatively low air speed. This radiator also 
permits the use of a large 48-in. diameter, 8-blade fan (Fig. 8) which 
can provide the required rate of air flow at a low top-speed of 740 
rpm. The fan is belt-driven from a sheave on a shaft extension at the 
commutator end of the generator. 


A further reduction of fan and air noise is obtained by the use of a 
variable-fan-blade-pitch control system which automatically controls 
the pitch of the blades in accordance with the cooling-water tempera- 
ture; thus no more air is passed through the radiator than is necessary. 
A manual control of the pitch of the fan blades permits the operator 
temporarily to feather the blades to zero pitch and stop the air flow, 
and the resultant noise, during a "take." 

The cooling air, after being drawn through the radiator, passes 
through the generator compartment and is exhausted through 
the muffler compartment door, and top louver-door whose opening 
can be adjusted. On many occasions the power plant can be oper- 
ated with the top louver-door closed, with all of the cooling air pass- 
ing through the muffler compartment and exhausted upward at the 
rear of the plant. All walls of the housing are insulated, as explained 
for the engine compartment, and therefore are deadened against vi- 
bration? caused by air flow. 

The carburetor-intake noise cannot be neglected. In this applica- 
tion outside air is drawn in through a baffled cap (Fig. 1), then 
through a large industrial-type air cleaner (Fig. 9), and into the car- 
buretor. The baffled cap and air cleaner function as silencers. 

5. Portability. The minimum over-all engine-generator set dimen- 
sions result from the following considerations: The 54-in. width ac- 
commodates a radiator having the desired capacity and in addition 
provides adequate clearance between the inside walls of the housing 
and the internal components for service and maintenance accessibility. 
The minimum height is determined on one end by a radiator of suf- 
ficient capacity, and on the other end by the combined heights of the 
engine and muffler. To reduce the latter, the engine mounting plat- 
form, consisting of steel channels welded crosswise and longitudinally, 
is recessed below the top surface of the main-base channel structure. 
Thus the minimum height requirement for the engine and muffler 
is made to conform to that required for the radiator. The length 
of the power plant is considerably reduced by the design of the 
short-coupled arrangement between the engine and the generator. 

Such engine-generator set cubicles are either carried on trucks or 
trailers. In the design of this plant, the 54-in. width of the enclosure 
is reduced by a curved contour near its base to 33% in., which is 
slightly less than the standard 34-in. over-all width of the main frame 
of commercial trucks having the capacity to carry this power plant. 
The 9J4-in. height above the bottom surface of the main base to the 
point where the power plant enclosure becomes 54 in. wide provides 
clearance over the rear wheels of the standard commercial trucks. 










Fig. 9. View through engine compartment service door on operator's side. 

This allows the power plant to be positioned low between the rear 
truck wheels, hence resulting in a minimum over-all height, and a 
minimum height of the center of gravity of the power plant above the 
ground. This same design feature is equally advantageous if the 
cubicle is to be mounted on a trailer. 

Arrangements are provided for handling the cubicle as a unit. 
Heavy steel tubular members are welded crosswise through the 
main-base frame near the front and rear and other tubular openings 
are located at both ends. These tubular openings are for insertion of 
heavy steel handling bars or wheel axles. 

6. Dependability. Dependability is necessary for trouble-free opera- 
tion, of all components. The engine chosen for this application is one 
which has proven its dependability in years of truck service, bus serv- 








Fig. 10. Bus bar panel. 

ice, in oil field locations and in other industrial applications. The 
generator is of a rugged design which has performed successfully in 
railway transportation equipment. The method used in coupling the 
generator to the engine insures long coupling life. 

Wherever practical, automatic operating components are supple- 
mented by manual controls to reduce the possibility of shut-down. 
The automatic voltage regulator is supplemented by manual field 
rheostat control. The pitch of the fan blades can be manually con- 
trolled in case of failure of automatic control. Should manual con- 
trol become inoperative, the blades can be mechanically locked in 
their pitched position and operation continued. 

There are, of course, inevitable possibilities of failure of equipment 
which could cause shut-down. The best insurance against such 







Fig. 11. Generator on dolly fixtures. 

occurrences is the use of components which have by service proven 
their reliability. 

7. Protective Devices. An overspeed governor, coupled to engine 
rotation, will open the ignition circuit and stop the engine before a 
damaging speed occurs. The engine cannot be restarted until the 
overspeed governor is manually reset by the operator. 

A safety switch in the ignition circuit will stop the engine if a loss 
of oil pressure occurs, or if the cooling-water temperature becomes too 

An overload relay, which will open the main line contactor, pro- 
tects the generator against overloads. 

The positive battery cable is grounded through a pin-plug arrange- 
ment so it can be conveniently and positively removed as a safety 


precaution during servicing, or when the power plant is idle. All bat- 
tery circuits, except the starter motor circuit, are fused. 

A Thyrite discharge resistor, permanently connected across the 
generator shunt field, protects the field insulation against high voltage 
breakdown by limiting the induced voltage which momentarily ap- 
pears when the field circuit is opened. 

The positive and negative bus bars for external feeder-cable lug 
connections (Fig. 10) are recessed behind an insulation panel, and an 
insulation barrier is located between them for protection against short 
circuits. The convenience outlets at the left of the bus bars are 

8. Maintenance. Door openings in the enclosure are such that all 
components are accessible for routine maintenance. The housing is 
constructed in sections to provide access for general overhaul. The 
top section, or muffler compartment, is removable as a unit. Each of 
the sides consists of two separate wall panels, and the rear is made 
up of one panel, all of which are individually removable. External 
stainless steel trim strips not only serve to cover the joints between 
sections for the sake of appearance, but also as mechanical members 
which tie the housing sections together. 

During a major overhaul of the power plant it may become neces- 
sary to detach the generator from the engine bell housing and separate 
the coupling. Adjustable dolly fixtures are provided (Fig. 11) which 
can be secured to the ears of the generator in place of the secondary 
supports. With the dolly wheels resting on the main base channels, 
the generator may be rolled in or out of engagement with the engine. 


In the making of motion pictures the production-time factor is of 
such importance that every precaution against possible power failure 
should be given the utmost consideration. The engine-generator set 
must be sufficiently rugged to withstand hard travel over rough roads, 
yet deliver the maximum of power with constant performance. In 
spite of the fact that it is a highly specialized piece of equipment, it 
should, so far as possible, be designed for, and constructed of, units 
which have been proven dependable in other fields. 


1. "Report of Studio-Lighting Committee," Jour. SMPE, vol. 51, pp. 431-436, 

Oct. 1948. 

2. "Report of Studio-Lighting Committee," Jour. SMPE, vol. 49, pp. 279-288, 

Sept. 1947. 

3. B. F. Miller, "A motion picture arc-lighting generator filter," Jour. SMPE, vol. 

41, pp. 367-373, Nov. 1943. 

Laboratory Practice 
Committee Report 


IN THE FALL OF 1948, the writer proposed to John A. Maurer, then 
Engineering Vice-President of the Society, that a committee on 
chemical engineering practice be organized to provide Society members 
with much needed information on chemical and chemical engineering 
practice as it applies to our industry. Mr. Maurer agreed with the 
proposal and suggested that such a committee be organized as a sub- 
committee of the Laboratory Practice Committee. A subcommittee 
composed largely of chemists and chemical engineers working in the 
motion picture industry was accordingly appointed. Later it was 
suggested that this subcommittee expand its activities to include the 
entire function of the parent Laboratory Practice Committee under 
its present chairman. The committee was organized with East 
Coast and West Coast Sections, the West Coast Section under the co- 
chairmanship of Vaughn Shaner. 

It was still intended that the parent committee would work pri- 
marily on chemical and chemical engineering problems as they relate 
to the motion picture industry. However, at the first meeting of the 
Committee held in New York City the members brought up so many 
other laboratory problems that the original objective concerning 
chemical problems was placed far down the agenda laid out at that 
meeting. Another meeting of the Committee was held in Hollywood 
during the Fall Convention of 1949 at which time the projects laid 
out by the East Coast Group were discussed and other projects added 
to the agenda. The complete proposed agenda came to be as follows : 

1. Design of a special leader for television films to replace the Academy theater 

2. Aid in the standardization of screen brightness for 16-mm projection. 

3. Establish a standard for the notching of 35-mm and 16-mm negative films. 

4. Investigate the possibility of modifying sound and picture reduction printers 
to print forward and backward, and to employ 2000-ft negative feed and take-up. 

5. Investigate the standardization of edge numbering of 16-mm. 

6. Study recommendations for the splicing of 16-mm films. 

7. Study recommendations regarding 16-mm projection emulsion position. 

8. Study methods of bringing data on chemical and chemical engineering 
developments to the attention of Society members. 

PRESENTED: April 28, 1950, at the SMPTE Convention in Chicago. 


214 JOHN G. STOTT August 

Other laboratory problems were discussed, but were tabled as either 
being under the jurisdiction of other Committees of the Society or 
entirely outside the sphere of activity of the Society. 

The third meeting was held in New York. At that meeting V. D. 
Armstrong was appointed as the Laboratory Practice Committee 
representative on the Films for Television Committee which is 
undertaking the design of a television leader; therefore Mr. Armstrong 
is functioning in an advisory capacity to that Committee on any 
television leader production or use problems as they concern the 
motion picture laboratory. A report on this leader is forthcoming 
from this Committee. 

Edward Cantor was appointed to make a cursory survey of 16-mm 
screen brightness presently used in New York, Midwest and Holly- 
wood laboratories for print examination. It was felt that this would 
represent a first move toward a recommended screen brightness of 
16-mm projection. When completed, these data will be submitted to 
the Screen Brightness Committee for their use in further studies. 

Also, at the third meeting the Chairman was instructed to write to 
manufacturers of sound and picture reduction printers suggesting 
that they design these printers for forward and backward printing 
and provide facilities for printing from 2000-ft negatives. At the 
present writing the response from the printer manufacturers indicates 
no particular problem on the 2000-ft negative feature, but opinion is 
unanimous that the forward and backward printing feature would 
entail expensive design changes. Further examination of this project 
will be made. 

Paul Kaufman was appointed to study recommendations for the 
notching of 35-mm and 16-mm negatives. This study has progressed 
satisfactorily, and a tentative standard for 35-mm will be proposed 
shortly. However, the 16-mm notching problem is going to be more 
difficult to resolve. It has been suggested by Lloyd Thompson that, 
in view of the present confusion and nonuniformity in 16-mm notch- 
ing, an entirely new scene exposure changing system be adopted as a 
standard. Mr. Thompson has suggested that a magnetic cue system 
be used, which will eliminate the need for an actual notch in the film. 
This proposal will be considered in subsequent meetings. 

The 40-frame edge numbering proposal for 16-mm was originally 
opposed because of the laboratory problem of picking replacements 
on 16-mm printed from 35-mm. When the proposed standard offer- 
ing the optional privilege of edge numbering at the 40-frame interval 
or not edge numbering at all was submitted to the Committee, this 
original objection was withdrawn. It was felt that 16-mm replace- 


ment from 35-mm preprint material could be picked by scene descrip- 
tion as has been done in the past. Realizing that a 16-mm edge 
numbering standard was badly needed, it was agreed that any further 
complication of the 16-mm edge numbering situation by the Com- 
mittee would be detrimental to the 16-mm industry. 

The Committee has gone on record as opposing any change in the 
present 16-mm standard for emulsion position during projection. 
The laboratories printing 16-mm from 35-mm preprint material, 
which is by far the greatest dollar volume of product, end up with film 
projecting according to the present standard. Any change in 16-mm 
projection standards would require extensive and expensive alterations 
in existing printers, etc. It was felt that a change in 16-mm pro- 
jection standards would impose severe economic hardship on the 
laboratory, greatly out of proportion to the benefit that would accrue 
to other film-using segments of the industry. 

Thus far, no work has been done in the splicing of 16-mm films. 

A fourth meeting of the Committee was held also in New York. 
At this meeting Mr. Cantor reported on the 16-mm screen brightness 
survey of New York laboratories. The survey of Midwest and Holly- 
wood laboratories is yet to be made. 

At this meeting the chemical and chemical engineering problems 
were again discussed. Irving Ewig was appointed to start some work 
on these problems. Mr. Ewig is to make a survey of chemical and 
chemical engineering magazines and journals having data of interest 
to our industry. The Committee has requested that the Society 
subscribe to several periodicals to be submitted to committee mem- 
bers for abstracting. The Committee has also requested of Clyde 
Keith, Editorial Vice-President, that JOURNAL space be made avail- 
able for publishing these abstracts to benefit all Society members. 
This was approved by Mr. Keith. Fred Bowditch, Engineering Vice- 
President, has authorized the reforming of a subcommittee on 
chemical engineering practice within the Laboratory Practice Com- 
mittee. Hence, the original organization plan of the Committee has 
been realized. 

A fifth meeting of the Committee was held at the 1950 Spring 
Convention in Chicago. At this meeting the afore-mentioned proj- 
ects were discussed briefly and a long discussion ensued with repre- 
sentatives of the Armed Services on the 16-mm projection screen 
brightness problem. Further help on this problem has been offered 
by the Armed Services in an effort to arrive at a suitable standard for 
16-mm screen brightness. 

68th Convention 

Feature items on the Papers Program include a Symposium on Film Registra- 
tion that will appeal particularly to designers of film handling equipment. Recent 
study of film perforation shape as it affects steadiness in the camera, registration in 
printing and projection life of release prints will be reported upon at length. 
The years of formal experience with film perforated to current standards, to- 
gether with surveys now under way, should provide a thorough engineering 
basis for proposed new film standards that will be of serious interest to all film, 
equipment and laboratory people. 

Most of one day will be devoted to papers on several application aspects of 
high-speed motion picture photography. Adequate time is to be allowed for dis- 
cussion from the floor of practical application problems. 

Editing magnetic sound tracks in motion picture studio production will high- 
light another session and tie in with papers on magnetic recording equipment and 
studio practices. 

Photographic sound recording on a new type of color motion picture release 
film will be discussed as will many other items of interest, including "T" stop 
calibration of camera lenses. 

the place is Lake Placid Club the time is October 16-20 reservations are 
now being accepted, so send the card you received recently (or write) to: Daniel 
Nelson, Reservations Manager, Lake Placid Club, Lake Placid, N. Y. families 
are more than welcome informality prevails you and your guests will enjoy 
the outdoor recreation. 

Members going to Lake Placid from the West should take the New York 
Central and change at Utica. Those from New York City, the South or areas 
connecting only with the Pennsylvania Railroad can arrange for overnight or 
day service from New York City directly to Lake Placid via the New York Cen- 

Air transportation from New York City can be made available on a charter 
basis provided there are enough reservations. Planes are tentatively scheduled 
to leave New York at 10 A.M. and 2 P.M. Sunday, October 15, and 10 A.M. Monday, 
October 16, with return flights leaving Lake Placid at 10 A.M., 2 P.M. and 6 P.M. 
Friday, October 20. Round-trip fare is $40.00. If you desire a reservation on 
the plane, your check must be received by Society Headquarters before September 
10. Please indicate your preference for departure tunes. 

Board of Governors 

On Wednesday, July 26, the Society's Board of Governors met for its third 
regular meeting in 1950. Fiscal affairs were discussed at length and the Board 
reports that, in general, operations for the first half-year compare very favorably 
with the budget estimate. One disappointing note, however, was the report 
on membership status which showed dues for nearly 10% of the entire member- 
ship still unpaid as of June 30. The Board is investigating the reasons for this 
heavy list of delinquents, so that appropriate steps can be taken to reinstate 
them, as well as to avoid a long list of delinquents next year. 

Candidates for Society offices selected by the Nominating Committee for the 
annual fall election were ratified by the Board as were nominees for Journal 


Award, Samuel L. Warner Memorial Award, SMPTE Progress Medal Award and 
Honorary memberships. 

Having been away from motion picture activities for a number of years, Louis 
Pacent had resigned from membership in the Society. He now plans to renew 
his interest in technical motion picture matters and his reinstatement as a 
Fellow received unanimous endorsement by the Board. 

Student chapters of University of Southern California and New York Uni- 
versity have both been active recently. To help them along the Board has 
appointed Loren L. Ryder and William H. Rivers as Society advisers. 

Engineering Committees 

The summer months in recent years have been periods of relative inactivity 
for the Society's engineering committees, but the current load of standards proj- 
ects and work related closely to television has kept several committees working 
diligently throughout the vacation season. Here, briefly, are two items of 

Theater Television 

Under the Chairmanship of D. E. Hyndman, the Theater Television Com- 
mittee's detailed study of performance requirements for interconnecting facilities 
will be continued with an examination of the effects of bandwidth variations, 
signal-noise ratio, distortion and compression on quality of the projected theater 
television picture. 

Representatives of the common carriers, as well as manufacturers of equipment 
for this new industry, have taken an active part and have highly praised the 
efforts of Otto Schade in his study of the four fundamental characteristics. Work 
on bandwidth requirements has progressed very well. A figure for admissible 
random noise has been proposed, based upon a detailed and objective sampling 
procedure developed by Mr. Schade, using the "noise" level of motion picture 
film as a reference. Subjective comparison of the experimental results between 
film and television pictures has already been made on a limited scale and will 
be repeated again, using full-scale commercial equipment for both pictures in 
the near future. 

Tentative conclusions have been made concerning square-wave distortion limits. 
Further work is now being done and will soon be discussed by the Committee. 

Screen Brightness 

For more than fifteen years, considerable time and effort have been devoted to 
the well-organized programs of the Screen Brightness Committee, work having 
begun seriously in the early thirties. A comparison method of estimating screen 
reflectivity was adopted, and sample gray cards to serve as reflection factor 
standards were bound into the JOURNAL for June 1933. Extended study of 
print density, vision and screen illumination produced a series of JOURNAL articles 
in the mid-thirties. 

Measurement methods have always been a serious problem. Just before the 
last war the Committee, under Frank Carlson, began to develop a photoelectric 
screen brightness meter, but the press of other urgent matters stopped the program 
shortly after a specification was agreed on. Three years ago the project was 
revived under the joint guidance of Erwin Geib and Bob Zavesky. A preliminary 


survey of eighteen theaters was encouraging and set the stage for an extended 
program by serving as a proving ground for several types of instruments, as well 
as providing a review of survey procedures. The Committee has now completed 
plans to start work on a somewhat larger survey of an estimated one hundred 
theaters ranging from small houses having fewer than five hundred seats to the 
largest in the country. Outdoor theaters and review rooms will also be included. 

The screen brightness meter recently developed for the Committee by Allen 
Stimson of General Electric has been doubly checked for accuracy and will be 
used in succession by six survey teams. Cities included and team leaders are: 
Los Angeles, C. W. Handley; Chicago, C. E. Heppberger; Toledo, A. J. Hatch, 
Jr. ; Rochester, F. J. Kolb, Jr. ; Philadelphia, C. R. Underbill, Jr. ; and New York, 
P. D. Ries. Considerable publicity for the survey has been given by the motion 
picture trade press, which will help to insure the co-operation of exhibitors and 
theater projectionists who were very generous with their time and assistance in the 
previous work. 

Letter to the Editor 

I was very interested to read in the March issue of the JOURNAL the article on 
spontaneous ignition of decomposing cellulose nitrate film and the appendix on p. 
381 on the film decomposition tests which have been carried out in this country 

There is, however, an error in the introduction to the latter, which I should be 
very grateful if you would correct. 

The British Film Institute is described as "a Government Department similar 
to the U.S. National Archives." Although the British Film Institute, including 
its National Film Library, is maintained chiefly by a grant from H.M. Treasury, 
it is not a Government Department in the full sense of the term. The only Gov- 
ernment Department concerned mainly with film preservation is that of the 
Government Cinematograph Adviser, at H.M. Stationery Office, which has in its 
care the films of the Imperial War Museum and those made by certain Govern- 
ment departments which are Crown copyright. 

The National Film Library of the British Film Institute is the only other official 
body in this country concerned with the permanent preservation of films and film 
records. Our scope, however, is wider in that we are concerned with the film, not 
only as an historical record, but also as an art, and the greater part of our collec- 
tion consists of nongovernment films. I imagine that whereas the Government 
Cinematograph Adviser's Department corresponds to the U.S. National Archives, 
the National Film Library here corresponds more nearly to the Library of Con- 
gress project which was hi operation some years ago. 

I hope that this clarifies the position. It is easy for confusion to arise because 
we co-operate most closely with the Government Cinematograph Adviser in all 
our preservation work and Mr. S. A. Ashmore, who advises the Government 
Cinematograph Adviser on technical matters, is also a member of our own Tech- 
nical Committee 


Deputy Director and Curator 

May 18, 1950 The British Film Institute 



H. G. Christensen, who was actively engaged in motion picture production 
since 1914, died in New York on June 10, at the age of 56. He was a native of 
Chicago and studied art before entering the field of training and sales-presenta- 
tion and film production. During World War I he was one of the Army's first 
instructors in aerial photography, having been assigned to the U.S. Army School 
of Aerial Photography at Rochester, N.Y. During World War II he directed 
the filming of many training subjects, including such secret ones as radar. He 
had been President of the West Coast Sound Studios and Vice-President of the 
Associated Sales Co., in charge of its motion picture department some years ago. 
He directed over 130 short subjects for Universal and had filmed many com- 
mercial and industrial subjects. He was co-author with L. S. Metcalfe of How 
to Use Talking Pictures in Business, published by Harper in 1938. 

Lauriston Everett Clark, 45, Director of Engineering for Technicolor Motion 
Picture Corp., died July 9, 1950, in Hollywood, Calif. Death was caused by a 
blood clot following an operation performed two weeks before. He was born 
at Haverhill, Mass., August 2, 1904, and attended the Massachusetts Institute 
of Technology. He served as an officer in the Chemical Warfare Reserve from 
1925 to 1926. Previously associated with Radio Corporation of America and 
Dunning-color, Mr. Clark joined Technicolor on January 1, 1943. He had been a 
member of this Society since 1929 and was also a member of the Academy of 
Motion Picture Arts and Sciences and the Optical Society of America. 

Journals Needed 

Motion picture technical books are being added to the cinema collection at 
Doheny Library, University of Southern California, and numerous issues of this 
Society's JOURNAL are needed to complete their reference file. Members who 
are willing to contribute any of those itemized below are asked to send a list of 
available copies to Boyce Nemec, Executive Secretary, so Society Headquarters 
can serve as a clearing house for all offers. In this way the Journals can be sent 
directly to USC without duplicating contributions or double handling of actual 

Under the Farmington Plan major libraries have been designated to collect 
publications from foreign countries on particular subjects, concentrating them at 
various locations within the United States. The University of Southern Cali- 
fornia has been designated the cinema center and consequently copies of every 
book on motion pictures published anywhere in the world will be sought for 
acquisition. Although official emphasis under this program is placed on books or 
pamphlets of foreign origin, USC also has a major interest in obtaining similar 
items which originate within the United States. The JOURNAL, being the in- 
dustry's major technical reference source, is an important item in this latter group. 

Issues currently missing from Doheny Library are : 

TRANSACTIONS: Nos. 16, 18, 30, 31, and 35. 

JOURNALS: Aug., Oct., Nov. and Dec. 1934; Mar., Sept. and Nov. 1938; 
Aug. and Dec. 1939; all months 1940; May and Aug. 1941; Apr. 1942; Sept. 
1943; all months 1944; Aug. 1945; Dec. 1946; all months 1947; all months 
1948; all months 1949; all months, to date, 1950. 


Central Section 

The National Electronics Conference in Chicago September 25-27, joined this 
year by the Institute of Radio Engineer's Chicago Section on the event of their 
25th Anniversary, will also have support of the Central Section of SMPTE. 
Technical Sessions of the three-day conference at Edgewater Beach Hotel include 
63 papers on a wide variety of subjects, ranging on the technical side from micro- 
wave spectroscopy, magnetic amplifiers and hermetically sealed ion chambers to 
a philosophical contribution "Is the Engineer Slipping" by E. A. McFaul, formerly 
of Northwestern University. Over 100 industrial exhibitors will be on hand with 
equipment displays. 

R. T. Van Niman, NEC Publicity Chairman and Past Chairman of the SMPTE 
Central Section, reports the NEC Television Session (10:00 A.M. Tuesday, Sep- 
tember 26) will be the September meeting of the Section. Three papers scheduled 

"Television in Industrial Applications," by J. A. Good, Diamond Power Specialty 

"Stereo Television in Remote Control," by H. R. Johnson, C. A. Hermanson and 

H. L. Hull, Argonne National Laboratory. 
"The Genlock: A New Tool for Better Programming in Television," by John H. 

Roe, RCA Victor Division. 

All SMPTE members are invited to attend this and any of the other 17 sessions 
and three luncheon addresses by: Wayne Coy, Chairman of the Federal Com- 
munications Commission; Mr. McFaul, noted above; and John V. L. Hogan, 
President, Interstate Broadcasting Co., Inc., and Past-President of I.R.E. 

Programs will soon be available from Mr. Van Niman. 

Book Reviews 

Film User Year Book, Volume II, 1950, edited by Bernard Dolman 

Published (1950) by Current Affairs Ltd., "Film User" Office, 19 Charing Cross 
Rd., London, W.C. 2. 320 pp., including advtg. 5% X %% in. Price, 10s. 

The Film User Year Book 1950 is a complete handbook of all the 16-mm and 
filmstrip activities in Great Britain for the year past. The various chapters 
deal with everything from projector placing diagrams and equipment to addresses 
to the British law in regard to the exhibitor. 

All 16-mm film produced in the sponsored (commercial), entertainment, 
scientific and industrial, church and armed service fields is not only listed, together 
with the distributors' names and addresses, but also cross-indexed as to title 
and subject. 

A complete census of film societies, distributors, producing companies, equip- 
ment manufacturers, recording studios, books and periodicals lists the names 
and addresses of practically everyone in the British Isles interested in any phase 
of motion pictures. 

It is quite interesting to note that many major producers in the United States 
are releasing 16-mm prints of late hits through their British offices. 

For the excellent data it contains, this book should be in the library of anyone 
interested in the British'film industry. WILLIAM K. AUGHENBAUGH, WLW-T, 
The Crosley Broadcasting Corp., Cincinnati, Ohio. 


The Organization of Industrial Scientific Research, by C. E. 
Kenneth Mees and John A. Leermakers 

Published (1950) by McGraw-Hill, 330 W. 42 St., New York 18. 368 pp. + 
15 pp. index. 20 figs. + 9 tables. 6 X 9 in. Price $5.00. 

The Organization of Industrial Scientific Research is both a guide and a stimulus 
to the clear thinking that is so necessary to the organization and operation of a 
successful research laboratory. A study of this book will aid management in 
making wise decisions regarding the need for research and the appropriations for 
carrying it on. It will help those engaged in the direction and administration 
of research to improve the effectiveness of their laboratories. It will open new 
avenues of thought and understanding to the scientist who is beginning to be 
interested in, or confronted with, administrative problems. 

While described as a "Revised Second Edition" (1st ed., 1920), this edition 
should be regarded as a new book with an excellent ancestor. 

Part I presents the very interesting history and background of organized 
research in order to provide for an understanding of its present position, or status, 
in government, universities and industry. The amazing growth of research 
is outlined, together with ideas regarding the future. 

Part II is concerned with existing research organizations. By example after 
example, by showing the needs, the accomplishments and the basic structures of 
organization, the present status of organized research is clearly presented. 

Part III comprises well-organized material bearing on the problems of research 
organization, administration and co-operation with other phases of industrial 
activity. Here is no handbook, but here is wisdom and inspiration to help all 
engaged in research to understand their environment and to solve the problems 
of organization of personnel and facilities. 

The authors have given freely of their knowledge and appreciation of the 
problems encountered in the organization of industrial scientific research, a field 
in which they are respected leaders. G. T. LORANCE, 125 Gates Ave., Montclair, 

New Members 

The following have been added to the Society's rolls since the list published last month. 
The designations of grades are the same as those in the 1950 MEMBERSHIP DIRECTORY: 

Honorary (H) Fellow (F) Active (M) Associate (A) Student (S) 

Askren, Lee T., Mechanical Engineer, Caldwell, Philip G., Engineer, American 
Eastman Kodak Co. Mail: 111 Broadcasting Co. Mail: 6285 Sunset 

Commodore Pkway, Rochester 10, Blvd., Hollywood, Calif. (M) 

N.Y. (M) Cooper, James B., Jr., Supervisor, Photo- 
Brown, Morris E., Supervising Design graphic Dept., Univ. of Michigan 
Engineer, Eastman Kodak Co. Mail: Aeronautical Research Center. Mail: 
48 Roosevelt Road, Rochester 18, N.Y. 603 Swift St " Ann Arbor, Mich. (A) 

(M) Craig, Bob, Distributor of Photographic 
_, _, Supplies, Craig Movie Supply Co. 

Brown, Robert A., Physicist, Remington Mail: 8414 Valley Circle, Canoga Pk., 

Arms Co., Inc., Physics Section, Calif (M) 

Bridgeport 2, Conn. (A) Culley ; p aul E>> Recording Eng i ne er, 
Buescher, R. E., Hollywood Sound Inst. Cinecraft Productions, Inc. Mail: 

Mail: 827 4th St., Apt. 102, Santa 2515 Franklin Ave., Cleveland 13, 

Monica, Calif. (S) Ohio. (A) 


D'Andrea, Matthew J., Free-Lance Tech- 
nician. Mail: 1589 Anderson Ave., 
Fort Lee, N.J. (M) 

Decker, Francis W., Motion Picture 
Technician, U.S. Air Force. Mail: 
109 Central Ave., Dayton 6, Ohio. (A) 

Didriksen, Roald W., Television XMTR 
Supervisor, KPIX Associated Broad- 
casters, Inc. Mail: 1056 Cole St., 
San Francisco 17, Calif. (A) 

Edous, John R., Univ. of Southern Calif. 
Mail: 3379 Santa Ana St., Huntington 
Pk., Calif. (S) 

Ettlinger, Adrian B., Electrical Engineer, 
Columbia Broadcasting System, Inc. 
Mail: 84-50 Fleet Court, Rego Park, 
L.I..N.Y. (A) 

Exner, William L., Television Engineer, 
KPIX Television. Mail: 2336-A 
Jones St., San Francisco 11, Calif. (A) 

Fouce, Frank, Motion Picture Producer, 
Theater Owner. Mail: 3212 Griffith 
Blvd., Los Angeles, Calif. (M) 

Innamorati, Libero, Engineer, Centre 
Sperimentale Cinematografia. Mail: 
Via Satrico, 43, Rome, Italy. (A) 

Jennings, James H., Chief Draftsman, 
Ampro Corp. Mail: 7006 N. Monon, 
Chicago, 111. (M) 

Johnson, Charles A., Motion Picture 
Camera Rentals, Mark Armistead, Inc. 
Mail: 911 N. Formosa Ave., Holly- 
wood, Calif. (A) 

Kagan, Phillip H., Service Manager, 
Du-Art Film Laboratories, Inc. Mail: 
845 Gerard Ave., Bronx 51, N.Y. (A) 

Kinney, Earl C., Manager, Film Process- 
ing Laboratory, Eastman Kodak Co. 
Mail: 1416 Sunset Terrace, Western 
Springs, 111. (M) 

Marino, Louis B., Hack Driver, National 
Transportation Corp. Mail: 911 East 
River Dr., New York, N.Y. (A) 

Martin, Gene F., Univ. of Calif., Los An- 
geles. Mail: 11024 Strathmore, Los 
Angeles 24, Calif. (A) 

Masters, Richard M., Tufts College En- 
gineering School. Mail: 124 Cotton 
St., Newton, Mass. (S) 

Maulfair, Robert J., Archer School of 
Photography. Mail: 946 Magnolia, 
Los Angeles 6, Calif. (S) 

Mochel, Walter E., Research Supervisor, 
E. I. du Pont de Nemours & Co., Inc. 
Wilmington, Del. (A) 

Muhl, Elinor P., Photographer, Dept. 
Aeronautical Engineering, Univ. of 
Minnesota. Mail: Rosemount Re- 
search Center, Bldg. 704-W, Rose- 
mount, Minn. (A) 

Niemann, Fred S., Motion Picture Pro- 
ducer. Mail: 942 Lake Shore Dr., 
Chicago 11, 111. (M) 

Niemeyer, John H., Industrial Technical 
Representative, Eastman Kodak Co. 

Mail: 204 Keotiak Rd., Park Forest, 
Chicago Heights, 111. (A) 

Pasquariello, Vincent J., Univ. of Calif., 
Los Angeles. Mail: 6208 Homes Ave., 
Los Angeles 1, Calif. (S) 

Risk, J. E., Chief Engineer, KSD-TV, 
The Pulitzer Publishing Co. KSD-TV. 
Mail: 351 Fairway Lane, Kirkwood, 
Mo. (A) 

Severdia, Anthony, Television Engineer- 
Projectionist, Associated Broadcasters, 
Inc., KPIX. Mail: 1440 Shafter Ave., 
San Francisco, Calif. (A) 

Snegoff, Mark, Teacher, Univ. of Calif., 
Los Angeles. Mail: 1954 Pinehurst 
Rd., Los Angeles 28, Calif. (A) 

Tunnicliffe, William W., Research Asso- 
ciate in Electronics (Airborne Tele- 
vision). Mail: 61-A Philips St., Water- 
town 72, Mass. (A) 

Verran, Bruce H., Univ. of Calif., Los 
Angeles. Mail: 2036 W. 79 St., Los 
Angeles 47, Calif. (S) 

Vinten, Charles, Managing Director, 
Messrs. W. Vinten, Ltd. Mail: 715, N. 
Circular Rd., London, England. (A) 

Wentzy, Woodrow P., Head, Dept. Audio- 
Visual Education and Photography, 
South Dakota State College, Brookings, 
S.D. (A) 

West, Lawrence, Television Camera 
Man, KPIX (Associated Broadcasters, 
Inc.). Mail: 1010 Curtis St., Albany 6, 
Calif. (A) 

Woods, George M., Projectionist-Repair- 
man, Grieme & Fasken Theatres 
Johnson's, Inc. Mail: Route 5, 
Wenatchee, Wash. (A) 

Yuen, Howard A., Radio Engineer, 
Associated Broadcasters, Inc. Mail: 
640 Second Ave., San Francisco 18, 
Calif. (A) 


Haines, Jesse H., Television Engineer, 
A. B. Du Mont Laboratories, Inc. 
Mail: 340 E. Olney Ave., Philadelphia 
0, Pa. (S) to (A) 

Newcombe, Charles R., In Charge, Elec- 
tronics, Hallen Corp., Bur bank, Calif. 
Mail: 1256 Elysian Park Ave., Los 
Angeles 26, Calif. (S) to (A) 


Payne, Raymond W. From: Laboratory 
Superintendent, Shelly Films, Ltd., 
Small Arms Administration Bldg., 
Longbranch, Toronto 14, Ont., 
Canada; To: Laboratory Superintend- 
ent, National Film Board of Canada, 
John & Sussex Sts:, Ottawa, Ont., 
Canada. Mail: 931 Bank St., Ottawa, 
Ont., Canada. 


New Products 

Further information concerning the material described below can be 
obtained by writing direct to the manufacturers. As in the case of 
technical papers, publication of these news items does not constitute 
endorsement of the manufacturer's statements nor of his products. 


Fastax High-Speed Motion Picture Cameras are now manufactured, sold and 
serviced by Wollensak Optical Co., Rochester, N. Y., as a result of a recent out- 
right purchase from the Western Electric Co. Fastax Cameras will be part of a 
newly formed division of Wollensak, known as the Technical and Industrial 
Division in High-Speed Photography. The new division is headed by John H. 
Waddell who had been with the Bell Telephone Laboratories since 1929. Under 
his guidance the Fastax was designed and perfected until it is called the world's 
fastest moving film high-speed camera and also the most versatile device for 
recording high-speed motion of either repetitious or transient nature. The first 
Fastax Camera to be designed used 16-mm film and took up to 4000 pictures a 
second. The second Fastax used 8-mm film at 8000 pictures a second. When 
a wider angle was desired for ballistics studies, a third Fastax was designed for 
35-mm half-frame wide-angle pictures. This camera gives 3500 pictures a second 
and has an angle view of 40 degrees or a width of field of 71 ft, when the camera is 
100 ft from the subject. 

A heat (infrared) deflector is described in Bulletin MI-318 available from Fish- 
Schurman Corp., 230 East 45th St., New York 17. It is a filter of multi-layer 
interference film and is marketed as the XUR-96 Heat Deflector. The manufac- 
turer reports that by installing the deflector at normal incidence in a motion pic- 
ture carbon arc projector, either of the condenser type utilizing 185 amp or of the 
reflector type using up to 125 amp, without additional blowers or other artificial 
cooling it was possible to reduce the heat on the film gate so that no buckling or 
embossing of the film occurred, and that the deflector is colorless and transmits 
between 94% and 96% of the visible light and reflects back to the source, depend- 
ing on what region of the infrared one measures, between 35% and 65%. 


The new G-E Flashtube No. 231 has 
been developed with opaque shields 
surrounding the tube's thoriated tung- 
sten electrodes designed to prevent 
formation of "incandescent trees," 
which in earlier flashtubes produced film 
travel "ghost," or double-image ef- 
fects on the screen. This new flash- 
tube is similar in principle to photo- 
graphic flashtubes developed by Gen- 
eral Electric during the war and since, 
and which are capable of emitting 
thousands of intense flashes of light 
with durations down to one-millionth 
of a second. This new flashtube for 
use as a television light source is in 
production and has a list price of $48.00 
plus tax. 

Meetings of Other Societies 

Sponsored by the Audio Engineering Society, the second Audio Fair will be held 
in New York City, October 26-28, at the Hotel New Yorker. There will be two 
floors of exhibits with demonstrations and technical papers scheduled for each 
of the three days. 

Illuminating Engineering Society, National Technical Conference, Aug. 21-25, 

Pasadena, Calif. 

Biological Photographic Assn., Annual Meeting, Sept. 6-8, Hotel Sheraton, 


Institute of Radio Engineers, West Coast Convention, Sept. 13-15, Long Beach, 


Institute of Radio Engineers, National Electronics Conference, Sept. 25-27, 


Theatre Equipment and Supply Manufacturers' Association, Annual Convention, 

Oct. 8-11, Stevens Hotel, Chicago 

Audio Engineering Society, National Convention, Oct. 26-28, Hotel New Yorker, 

New York 

Optical Society of America, Oct. 26-28, New York 

Theatre Owners of America, Annual Convention, Oct. 30-Nov. 2, Shamrock 

Hotel, Houston, Texas 
Acoustical Society of America, Fall Meeting, Nov. 9-11, Boston 

SMPTE Officers and Committees: The roster of Society Officers 
was published in the May JOURNAL. The Committee Chairmen and 
Members were shown in the April JOURNAL, pp. 515-22; changes in 
this listing will be shown in the September JOURNAL. 


Journal of the Society of 

Motion Picture and Television Engineers 


New Television Camera Tubes and Some Applications Outside the 

Broadcasting Field V. K. ZWORYKIN 227 

CBS Television Staging and Lighting Practices. . . RICHARD S. O'BRIEN 243 

Motion Picture Instruction in Colleges and Universities 

Synchronous Recording on l /rln. Magnetic Tape . . WALTER T. SELSTED 279 

Electrical and Radiation Characteristics of Flashlamps 

H. N. OLSEN and W. S. HUXFORD 285 
The Cine Flash A New Lighting Equipment for High-Speed Cinepho- 

tography and Studio Effects. . . .H. K. BOURNE and E. J. G. BEESON 299 

A New Heavy-Duty Professional Theater Projector 


A New Deluxe 35-Mm Motion Picture Projector Mechanism 

H. J. BENHAM and R. H. HEACOCK 319 

68th Convention 327 

Engineering Committees Activities. 327 

High-Speed Photography Question Box 328 


The American Annual of Photography, Volume 64, edited by Frank R. Fraprie 

and Franklin I. Jordan Reviewed by John W. Boyle 331 

Practical Television Engineering, by Scott Helt 

Reviewed by E. Arthur Hungerford, Jr. 331 

Sound Absorbing Materials, by C. Zwikker and C. W. Kosten 

Reviewed by Hale J. Sabine 332 
American Cinematographer Hand Book and Reference Guide, Seventh Edition, 

by Jackson J. Rose Reviewed by John W. Boyle 333 

Theatre Catalog, 8th Annual Edition Reviewed by Leonard Satz 333 

Current Literature 334 

New Members 335 

New Products 336 



Chairman Editor Chairman 

Board of Editors Papers Committee 

Subscription to nonmembers, $12.50 per annum; to members, $6.25 per annum, included in 
their annual membership dues; single copies, $1.50. Order from the Society's General Office. 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and 
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture and Tele- 
vision Engineers, Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General 
and Editorial Office, 342 Madison Ave., New York 17, N.Y. Entered as second-class matter 
January 15, 1930, at the Post Office at Easton, Pa., under the Act of March 3, 1879. 
Copyright, 1950, by the Society of Motion Picture and Television Engineers, Inc. Permission 
to republish material from the JOURNAL must be obtained in writing from the General Office of 
the Society. Copyright under International Copyright Convention and Pan-American Con- 
vention. The Society is not responsible for statements of authors or contributors. 

Society of 

Motion Picture and Television Engineers 

342 MADISON AVENUE, NEW YORK 17, N.Y., TEL. Mu 2-2185 

Earl I. Sponable 
460 W. 54 St. 
New York 19, N.Y. 

Peter Mole 

941 N. Sycamore Ave. 
Hollywood 38, Calif. 



Loren L. Ryder 
5451 Marathon St. 
Hollywood 38, Calif. 

Clyde R. Keith 
120 Broadway 
New York 5, N.Y. 

Robert M. Corbin 
343 State St. 
Rochester 4, N.Y. 

William C. Kunzmann 
Box 6087 
Cleveland 1, Ohio 

Fred T. Bowditch 
Box 6087 
Cleveland 1, Ohio 


Frank E. Cahill, Jr. 
321 W. 44 St. 
New York 18, N.Y. 

Ralph B. Austrian 
25 W. 54 St. 
New York 19, N.Y. 


Herbert Barnett 
Manville Lane 
Pleasantville, N.Y. 

Kenneth F. Morgan 
6601 Romaine St. 
Los Angeles 38, Calif. 

Norwood L. Simmons 
6706 Santa Monica Blvd. 
Hollywood 38, Calif. 


Charles R. Daily 
113 N. Laurel Ave. 
Los Angeles 36, Calif. 

John P. Livadary 
4034 Cromwell Ave. 
Los Angeles, Calif. 

William B. Lodge 
485 Madison Ave. 
New York 22, N.Y. 


Lorin D. Grignon 
1427 Warnall Ave. 
Los Angeles 24, Calif. 

Paul J. Larsen 

4313 Center St. 
Chevy Chase, Md. 

William H. Rivers 
342 Madison Ave. 
New York 17, N.Y. 


F. E. Carlson 
Nela Park 
Cleveland 12, Ohio 

George W. Colburn 
164 N. Wacker Dr. 
Chicago 6, 111. 

Edmund A. Bertram 
850 Tenth Ave. 
New York 19, N.Y. 

Malcolm G. Townsley 
7100 McCormick Rd. 
Chicago 45, 111. 

Edward S. Seeley 
161 Sixth Ave. 
New York 13, N.Y. 

R. T. Van Niman 
4501 Washington St. 
Chicago 24, 111. 

New Television Camera Tubes 
And Some Applications 
Outside the Broadcasting Field 



SUMMARY: The operation and performance characteristics of television 
camera tubes from the iconoscope to the image orthicon and vidicon are 
described briefly, stressing recent developments. The application of the 
vidicon in industrial television equipment and, in greater detail, possible 
uses of television techniques in astronomy are outlined. 

THE DESCRIPTION of the television camera as an electric eye is as 
old as television itself. The reason for this is obvious, in view 
of the function of the television camera. Yet, to anyone viewing 
the images produced by early television systems, the term may well 
have seemed presumptuous and the comparison with the human eye 

The human eye is indeed a marvelous mechanism. It functions 
efficiently over a brightness range of one to one-hundred million, 
10 ~ 6 foot-Lamberts to 10 2 foot-Lamberts. At low light levels 
quanta incident over a cone with a vertex angle of 30 min co-operate 
to produce a single visual sensation, while at high light levels visual 
angles as small as one minute are resolved. With a quantum ef- 
ficiency of the order of 5% at low light levels, the contrast recognition 
of the eye is limited only by the statistical fluctuation in the in- 
cidence of the light quanta. This is given simply by the square root 
of the total number of quanta imaging a "picture element" of the 
object on the retina within the storage period of the eye, approxi- 
mately a fifth of a second. It has been found experimentally that the 
recognition of brightness contrast between two picture elements of 
equal size requires that their average brightness difference should 
exceed the statistical fluctuation in brightness at least by a factor 
of 5. If, for instance, 1000 quanta on the average are absorbed in a 
particular picture element projected on the retina within the storage 
period of a fifth of a second, the statistical fluctuation will be 30 
quanta or 3% of the element brightness; thus, in order that an ele- 
ment of equal size be distinguishable from the first element, it must 
PRESENTED: October 13, 1949, at the SMPE Convention in Hollywood. 





differ from it by 15% in brightness. This is the threshold contrast 
under the conditions here described. It is seen that, in general, the 
threshold contrast is inversely proportional to the square root of the 
brightness and the picture element area. 

The same principles which govern the contrast recognition and 
sensitivity of the eye apply to any other light-sensitive devices, such as 
television camera tubes. The fact that photosensitive surfaces 
with quantum efficiencies of 5% and better are available suggests 
that, ultimately, pickup devices with sensitivity equal to that of the 
human eye can be achieved. Today this goal has in fact been at- 
tained, at least within limited ranges of operation. A brief outline 



Fig. 1. Diagram of the iconoscope. 

of the development will indicate the nature of the obstacles which 
had to be overcome and show where further progress may be ex- 
pected in the future. 

The first television camera tube which shows a close correspondence 
with the properties of the eye, the iconoscope, was conceived some 26 
years ago. 

In detail, the iconoscope (Fig. 1) functions in the following manner: 
The picture is projected on a mosaic of minute sensitized silver 
globules deposited on a mica plate. Under the influence of illumina- 
tion these globules give off photoelectrons and are charged positively; 
the change in potentiaMs determined both by the charge lost and the 
capacity of the element relative to a metal film on the back of the 
mica, the signal plate. 

The surface of the mosaic is scanned in a regular line pattern 


525 lines every thirtieth of a second by a sharply focused electron 
beam. The beam restores the original electrical potential wherever 
it strikes the mosaic. In doing so it releases from the signal plate a 
charge equal to the charge stored in the scanned element and this 
released charge applies a voltage pulse proportional to the stored 
photoelectric charge to the picture amplifier. The succession of 
these voltage pulses constitutes the video signal. Applied to the 
grid of a viewing tube., whose fluorescent screen is scanned in unison 
with the iconoscope mosaic, it reconstructs on the viewing screen the 
image projected on the mosaic. 

A vital property which the iconoscope shares with the human eye 
and which was absent from all earlier television systems is storage. 
The video signal of the iconoscope is determined not only by the 
photoelectric charge released by the light at the instant of scanning, 
as in nonstorage systems, but by the charge stored by the light in the 
picture element considered throughout the period between successive 
scannings. Thus storage permits an increase in sensitivity by a 
factor equal to the total number of picture elements. 

This gain is not realized in full in the iconoscope. The field condi- 
tions in front of the mosaic prevent efficient storage of charge and 
efficient utilization of stored charge for the formation of the video 
signal. These and other secondary factors combine to make the 
sensitivity of the iconoscope less than that of the eye by 3 to 4 orders 
of magnitude at low light levels. 

In the iconoscope the scanning beam consists of 1000-v electrons 
which, at incidence on the mosaic, eject a considerable number 
of secondary electrons for every incident beam electron. The 
"redistributed" secondary electrons give rise to a spurious signal or 
"shading," which may require manual compensation by the monitor- 
ing engineer. The weak retarding field in front of the mosaic also 
serves to return to the emitting element, or to redistribute, a large 
fraction of the photoelectrons, impairing the efficiency of charge 
storage. Under normal operating conditions these causes reduce 
the video signal amplitude to about 5% of the value attainable with 
ideal collection. 

It should be noted that the same factors which lead to relative 
inefficiency and the presence of shading also have some desirable 
consequences. These are perfect stability at all Jevels of illumina- 
tion and a nonlinearity of response which compensates the non- 
linearity of opposite sign of the viewing tube. 

The first successful improvement on the iconoscope consisted in 
projecting the light image on a continuous transparent photocathode 




and employing the photoelectrons so generated to form an electron 
image on the mosaic. The charges stored in this manner on the 
mosaic, and hence the output signal, were greater than in the ordinary 
iconoscope for two reasons: The continuous photocathode was more 
photosensitive than the mosaic and five or more secondary electrons 
left the mosaic for every primary photoelectron incident on it. The 





Pig. 2. Diagram of image orthicon. 

Fig. 3. Image orthicon. 

"image iconoscope," formed in this manner was, in fact, up to ten 
times more sensitive than the ordinary iconoscope. Modern versions 
of it are employed today, with considerable success, in France. 

The image iconoscope increased the signal level and hence the 
sensitivity of the pickup tube, but left the unfavorable field condi- 
tions in front of the mosaic essentially unaltered. The orthicon on 
the other hand is distinguished by the presence of a strong collecting 


field in front of the mosaic which guarantees complete collection of 
photoelectrons and secondary electrons and prevents the occurrence 
of spurious shading signals. This occurs when the surface is approxi- 
mately at cathode potential. 

The difficult problem which had to be solved before the orthicon 
became practical consisted in attaining a very small spot size at low 
electron velocities and retaining this, as well as the perpendicular 
direction of incidence of the beam on the mosaic, for all deflections. 
The problem was finally solved by immersing the tube in a longi- 
tudinal magnetic focusing field and superposing horizontal and 
vertical magnetic deflecting fields on it. The secondary electrons 
as well as those which are turned back in front of the mosaic follow 
very nearly the same paths as the incident beam and are ultimately 
collected by a diaphragm in front of the cathode. This method of 
focusing has proved so successful that better than thousand-line 
resolution has been achieved with it on a mosaic approximately two 
inches in height. 

The performance of the orthicon is as expected. There is a com- 
plete absence of shading signals and a strictly linear relation between 
signal and light up to a certain limit. Its sensitivity is approxi- 
mately an order of magnitude greater than that of the iconoscope. 
It has given good service and in modified form is finding wide em- 
ployment in England today. 

The perfect stability of the iconoscope, the highly sensitive con- 
tinuous photocathode of the image iconoscope, the freedom from 
shading of the image orthicon, and a high level of signal output made 
possible by secondary emission multiplication were finally combined 
in the image orthicon, developed in the middle forties. This in- 
genious tube is represented schematically in Fig. 2. The two-sided 
target, which takes the place of the mosaic, consists here of a very 
thin film of glass, whose conductivity is high enough so that differences 
of potential on its two opposite faces are largely wiped out by con- 
duction in the course of a frame time. Yet the film is so thin that 
charge leakage from one picture element to its neighbors, within the 
same period, is negligible. To the right of the target is the image 
section of the tube. Photoelectrons ejected from the photocathode 
by the light image of the scene to be transmitted are focused magneti- 
cally on the glass target. Here they eject secondary electrons, which 
are either collected by a very fine-meshed target screen in front of the 
target or turned back to the target. 

To the left of the target is the scanning section of the tube, which 
in most details resembles that of the orthicon already described. 




However, the signal is carried by the return beam, from which a 
number of electrons equal to those removed from the target by 
secondary emission are abstracted. The return beam is incident on a 
diaphragm surrounding the scanning beam and ejects from it second- 
ary electrons which spill over into an array of secondary emission 
sensitized pinwheels surrounding the cathode. These pinwheels 
function as stages of a secondary emission multiplier which amplifies 
the return beam by a factor of 300 to 1000. In the picture of the 
image orthicon (Fig. 3) the pinwheels are visible near the base of the 

The variation of the signal strength with illumination on the 
photocathode is shown for three image orthicons in the next figure 



f 6 


2 6 
1 2 











r ~ 













_- 2 

< b 
O 4 


' 7 















4 6 

n 246 

2 46 

2 46 

0.0001 0.001 001 


Fig. 4. Response characteristics of three image orthicons. 


(Fig. 4). As long as the target potential remains below that of the 
target screen, the collection of the secondary electrons is virtually 
complete and the response perfectly linear, as indicated by the first 
part of the two curves. On the other hand, for higher intensities 
of illumination the response rapidly levels off and soon ceases to in- 
crease with further increase in illumination. Redistribution of 
electrons at the boundaries between regions corresponding to different 
illuminations then serves to establish edge contrasts which result in a 
fairly natural rendition of the scene. 

It has been shown by Dr. Rose of the RCA Laboratories that the 
performance of any pickup device can be adequately described 




in terms of the brightness, contrast and angular dimensions of detail 
of the object that can be perceived with the device employing a lens 
of given aperture and a given storage time. These several factors 
combine to yield a performance figure which, for an ideal system, 
becomes equal to the quantum efficiency of the primary photoprocess 
taking place in the device. Figure 5 shows the performance param- 
eter as function of scene brightness for the human eye, motion 
picture film and several television pickup tubes. As might have been 
expected, no man-made device approaches the human eye in its 


Fig. 5. Figure of merit of human eye, motion picture film 
and television pickup tubes as function of scene brightness (A. Rose). 

range of satisfactory performance. Among man-made devices, on 
the other hand, the great superiority of the image orthicon at low light 
levels is clearly evident. It is seen, in fact, that the image orthicon 
has within a considerable range a performance figure of the order 
of 1%, that is nearly equal to the quantum efficiency of its photo- 
cathode; hence in this range it can justly be regarded as an ideal 

It would be mistaken, however, to consider the image orthicon as 
an endpoint in camera tube development. Apart from the obvious 
goal of greater sensitivity of the photocathode and extension of the 




range to still lower light levels, there is the challenge of simplifying the 
tube and its auxiliary equipment. It is clear that a tube of the 
complexity of the image orthicon presents many difficult production 
problems. In addition, added power supplies are needed for the 
secondary emission multiplier and the image section and the corre- 
spondingly large number of controls complicate the adjustment and 
servicing of the image orthicon camera. 

A considerable advance in the properties of the photocathode has 
been achieved by substituting for the red-sensitive silver-cesium oxide 
(2P23, Fig. 6) and the blue-sensitive antimony-silver cesium surfaces 




2000 4000 6000 8000 10000 12000 14000 16000 


Fig. 6. Spectral sensitivities of different types of image-orthicon photocathode.". 

(5655 and 5769, Fig. 6) a new type of bismuth-silver cesium surface 
(5820, Fig. 6) . This not only doubles or triples the sensitivity of the 
earlier tubes, but greatly improves the match to the color sensitivity 
of the eye, leading to a more faithful and pleasing reproduction of the 
transmitted picture. The curves in Fig. 6 show the relative spectral 
sensitivites of the three types of photosensitive surfaces. 

A great simplification in the pickup tube construction and the 
auxiliary equipment, without corresponding loss in sensitivity, has 
recently been attained in a new type of pickup tube, the vidicon. 

Through the provision of a suitable photoconductive target de- 
posited on a transparent signal plate, the intrinsic sensitivity of the 








Fig. 7. Diagram of vidicon. 

tube has here been raised to such a level that it has become permissible 
to dispense with the image section and secondary-emission multi- 
plier. Low-velocity scanning is employed as in the .orthicon and 
image orthicon: Light-induced conduction causes, in view of the 
positive bias on the signal plate, illuminated portions of the target 
to become positive. The positive charge so stored is neutralized 
by the scanning beam, giving rise to picture signal pulses in the 
signal plate lead (Fig. 7) . 

The small dimensions of the vidicon (Fig. 8) it is only 1 in. in 
diameter fit it ideally for industrial television purposes, where 

Fig. 8. Comparison of vidicon and image orthicou 




Fig. 9. Industrial television system incorporating vidicon camera. 


Fig. 10. Stereo television camera with control unit and monitor. 




compactness and portability are decisive advantages. Figure 9 
shows a vidicon camera weighing only 8J/2 lb, together with a com- 
plete monitor control unit weighing some 50 Ib. In practice, the 
control unit may be located 500 ft from the camera, being connected 
to it by cable. Equipment of this type has found many applications 
in industry, for surveillance and research, and in education, for the 
demonstration of microslides and surgical operations. For the latter 
purpose the three-dimensional representation yielded by a stereo- 
scopic television camera (Fig. 10) proves particularly valuable. 

The great sensitivity of the newer tubes which have just been 
described makes them eminently suitable for the transmission of 
pictures in natural color. High sensitivity is here needed, since the 
process of separating out the primary component colors of the picture 
invariably leads to a considerable loss of intensity. One system of 
color transmission which possesses the advantage of being readily 

Fig. 11. Optics of image orthicon color camera for direct pickup. 

fitted into the existing system of black-and-white broadcast television 
records the three primary color pictures on separate pickup tubes. 
Figure 11 shows schematically the arrangement of an appropriate 
color pickup camera employing three image orthicons. Dichroic 
reflectors serve to separate the light from the scene into its primary 
color components, so that different-color images, of identical size, 
are formed on the photocathodes of the three tubes. 

The primary aim in the development of the devices which we have 
considered so far has been the creation of tools for a satisfactory 
broadcast television service. Yet their usefulness, and the usefulness 
of apparatus which may be readily derived from them, goes far beyond 
this, as we have already seen in connection with the industrial tele- 
vision camera. In particular in the scientific field television tech- 
niques can often be applied to great advantage. It is true that 
the requirements of science and entertainment are so different, often 




Light from 

Top view cf light chopper 


Fig. 12. Photoelectric clock drive 
Correction (Whitford and Kron). 

Fig. 13. The Telectroscope. 


even diametrically opposed to each other, that our attitudes and 
methods in the two fields must need be quite different. We shall 
indicate some ways in which television methods may find application 
in the field of astronomy. 

An obvious use is to let the television camera substitute for the 
observer at the eyepiece of the telescope, making possible remote 
control of the instrument with a minimum of thermal and other dis- 
turbances. Even if the astronomer himself might not deem it ad- 
visable to separate himself to that extent from his telescope, he might 
readily appreciate the advantages of letting visitors view his equip- 
ment and the stars with television eyes instead of with their own. 

An electronic technique derived from television development may 
also be employed to flatten the image field of the Schmidt Camera 
and, eventually, increase its sensitivity. To this end the curved 
cathode of an image tube is made to coincide with the focal surface 
of the Schmidt Camera. This photocathode is imaged electronic- 
ally on a flat fluorescent screen. In order to photograph the si- 
dereal image, a photographic plate is placed in contact with the 
screen. Since the number of quanta ejected from the fluorescent 
screen by each accelerated electron may be made to exceed con- 
siderably the average number of quanta required to free a photo- 
electron from the photocathode, a shorter exposure time will be 
needed to leave a visible star record on the photographic plate. It 
is true that, though the image produced at the fluorescent screen is 
brighter than that at the photocathode, it is also "noisier," that is, 
contains less intrinsic information. 

Television techniques may, furthermore, be employed to advantage 
for stabilizing star images. Whitford and Kron at Washburn Ob- 
servatory many years ago installed a photoelectric guiding mech- 
anism on a telescope to correct the clock drive (Fig. 12) . A selected 
fixed star is imaged on the edge of a roof prism, which directs the 
split beams through opposite sides of a rotating 180 sector disk 
onto a multiplier phototube. If the intensity of the two beams is 
unequal an alternating current is generated which is employed to 
correct the clock drive so that the star image remains centered on the 
edge of the roof prism. Some time ago the author proposed an all- 
electronic system with the corresponding almost complete absence of 
inertia for compensating fluctuations in atmospheric refraction 
(Fig. 13). In the figure, the sidereal image is formed on the photo- 
cathode of an image iconoscope, the electron image of a particular- 
fixed Star being centered on a small aperture in the middle of the 
mosaic. The electrons forming this image fall on the vertex of a 




pyramid, whose four sides act as first dynode for four electron multi- 
plier structures. The output currents of the four multipliers pass 
through four deflecting coils, which serve to maintain the fixed star 
image centered on the pyramid, so that the sidereal image reproduced 
on a viewing tube screen appears stationary. 

The above method requires the construction of a highly specialized 
complex electronic centering tube. The same end may, however, 
be accomplished with the aid of conventional television equipment 
in conjunction with suitable gating circuits. Figure 14 shows a block 
diagram of the circuit which may be employed for this purpose. 






In _J______- 


! i 
















Fig. 14. Electronic image stabilizer for sidereal images. 

The sidereal image is projected by the telescope on the photosensitive 
surface of the pickup tube so that the fixed star serving as "guiding 
star" is centered on it. Apart from the amplifiers and the television 
receiver on which the sidereal image is reproduced, centering circuits 
are provided whose purpose it is to maintain the scanning pattern 
centered on the guiding star image, independent of atmospheric 
fluctuations and minor clock errors. Here the output signal of the 
pickup tube is applied to two gating circuits which are sensitized 
over a series of time intervals covering a region of a few line widths 
about the prescribed position of the image of the guiding star in the 
scanning pattern. If the star image moves slightly in the course of 




a frame time, the signal pulse causes the horizontal gating circuit 
to apply a horizontal centering signal to the tube deflection. The 
same pulse causes the vertical gating circuit to apply a vertical center- 
ing impulse corresponding to the difference in the line number of the 
actual and prescribed occurrence of the star pulse. To take account 
of storage, the gating circuits are blanked by the first impulse if, and 
only if, the latter occurs before the prescribed occurrence of the star 

In speaking of the electronic stabilization of star images to com- 
pensate atmospheric disturbances and clock drive errors a subject 
discussed last year by Professor Zwicky in Zurich we have been able 
to keep our feet on the ground. We shall now contemplate the 
possibility of placing our telescope in a balloon and ascending to 










Fig. 15. Image stabilizer for balloon-mounted Schmidt telescope. 

heights where atmospheric refraction ceases to play an appreciable 
role. We will now need electronic methods to maintain the proper 
orientation of the telescope in spite of the uncontrollable motions of the 
balloon and, eventually, to send the star images back to earth. Even 
though this method would not be applicable to the largest telescopes, 
the possibility of almost completely eliminating atmospheric effects 
might well render it of value. 

We shall again direct our attention to the crucial problem, the 
stabilization of the sidereal image. As compared with the system for 
compensating atmospheric fluctuations, we must now count with the 
possibilities of much greater total displacements and of the 'rotation 
of the image. We can take account of the first factor by providing 
servomotors for the rotation of the telscope about a vertical axis and 
about a tilt axis (Fig. 15). These servomotors are controlled by the 
centering currents for the scanning pattern. A third servomotor is 



provided to rotate the deflection yoke about the tube axis. This 
motor is controlled by pulses from an auxiliary guiding star, normally 
located near the periphery of the scanning pattern, on the horizontal 
line through the guiding star image. The circuit arrangement for 
this system is indicated in Fig. 16. 


Fig. 16. Block diagram 
of image stabilizer. 

We have mentioned a few ways in which television pickup tubes 
may perform a service in the field of astronomy. There are without 
doubt many other ways whose discovery demands familiarity with 
both the problems of astronomy and the possibilities and limitations 
of electronic equipment. Experience has shown consistently that 
material progress in any one field of science and engineering has had a 
beneficent effect on the development of all other fields. The de- 
velopment of electronic pickup tubes with sensitivities of the same 
order as that of the human eye should be no exception. 

CBS Television Staging 
And Lighting Practices 



SUMMARY: Television as a visual medium must be operated within the 
boundaries of its technical characteristics to achieve good visual reproduction. 
The system handles a limited range of luminance, introduces luminance 
transfer distortion, exhibits spurious effects (halos, image orthicon ghosts, 
clouding, streaking) and has finite detail resolution. It is necessary to pro- 
vide guidance whereby production personnel can fully exploit the present 

Accordingly, rules have been formulated for each of the major production 
operations at CBS, viz., staging, lighting, camera operation and direction. 
Individuals working in these phases are thus enabled to perform their sepa- 
rate functions with assurance that their combined efforts will produce images 
which are both technically correct and artistically pleasing. 

IF TELEVISION were a perfect visual reproduction medium, it would 
be possible to allow qualified artistic judgment to be the sole 
arbiter of staging practices. Television is not, as yet, a perfect 
transmission system. At the present stage of technical development 
it is necessary to temper artistry with technicality to respect rules 
which recognize the characteristics of the present facilities. 

It is an important engineering function to work toward improve- 
ment of technical performance. While this work is in progress, it is 
equally important to study the equipment as it exists and to determine 
the necessary boundary conditions in order that rational artistic- 
technical compromises can be made in current program production. 
The television studio practices which are discussed here have been 
found helpful in day-to-day operations by the Columbia Broadcasting 

These practices, concerned principally with the control of scene 
luminance and content, are outlined in groups of co-ordinated rules 
for use in the staging (scenery preparation), lighting adjustment and 
camera operations phases of production. The work of the various 
production departments, though separated in time and location, is 
thereby guided to obtain picture quality which avoids known pitfalls 
and makes the best possible use of the television system. 

In this presentation, a review of certain technical characteristics of 
the present facilities, including illustrations of spurious effects which 

PRESENTED: April 25, 1950, at the SMPTE Convention in Chicago. 





Fig. l-A. Halo. 

Fig. 1-B. Halo. 


Fig. 1. Halo 

The heavy black fringe surrounding the young lady in A results from a shower 
of low-velocity secondary electrons emitted from the high-light areas on the image 
orthicon target. These electrons land on the surrounding dark area completely 
discharging the nearby portions. In B, a lighter background has been substituted 
and the halo has been greatly reduced. In this case, the secondary emission from 
the background area is now sufficient to alter the field configuration in the vicinity 
of the target causing the excess high-light area electrons to land properly on the 
collector mesh. 

NOTE: The dashed-line ellipse is a mask placed on the picture monitors to 
bound the area within which essential picture information should be held to 
prevent subsequent cropping by receiver masks or the film recording process. 

occur, is followed by the statement and discussion of some of the more 
important working rules which have been formulated. The intention 
is to indicate the nature of the approach which has been made toward 
control of production practices. 


Aspects of the present-day television facilities which may constitute 
technical limitations are : total contrast range, shape of the contrast 
gradient or luminance transfer characteristic, interaction among adja- 
cent picture areas, and detail resolving capability. The characteris- 
tics of the image orthicon pickup tubes, of the electrical transmission 
system and of the reproducing cathode-ray tubes all may contribute to 
the over-all distortion in picture quality. For a network originating 
studio, it is particularly important that careful control be exercised as 
network transmission facilities and television film recordings used for 
program distribution certainly cannot be expected to minimize picture 
defects produced in the studio. 

Limited total contrast range constitutes one of the most basic 
problems. The ranges of luminance values which can be handled by 
several familiar systems are approximately as follows : 

The human eye for a particular luminance adaptation 100 to 1 

35-Mm motion pictures typical projection. ... 40 to 1 

Live, direct-view television ideal conditions 40 to 1 

Live, direct-view television typical conditions ... 20 to 1 

Kodachrome-type color film processes 15 to 1 

It is important that, in the television scene to be transmitted, all 
subject matter have a luminance within the approximate 20-to-l 
range handled by the system. 

Distortion in luminance transfer characteristics may have the 
effect of increasing the apparent contrast between areas of a scene. 



Fig. 2-A. Image orthicon ghost. 

Fig. 2-B. Image orthicon ghost. 




Fig. 2. Image Orthicon Ghost 

The displaced image of the hand shown in A results from high-velocity second- 
ary electrons emitted from the high-light area on the image orthicon target. These 
electrons travel as a group back toward the photo-cathode, eventually deceler- 
ating and returning to the target, producing a positive signal by knocking off 
secondary electrons as in the case of an ordinary electron image signal. The 
displacement results from travel through the axial magnetic field present. Here, 
too, it is possible to minimize the landing of this ghost image by maintaining less 
contrast between the high-light and the background areas. However, in B, the 
hand has simply been moved to the center of the raster, where the ghost travels 
along the axis of the tube and falls back on the high-light area without displace- 
ment. Slight defocusing of image focus would also reduce the ghost but at a 
sacrifice in resolution. 

Through a typical direct-view receiver, for example, a 2-to-l scene 
brightness difference may under some conditions appear as a 6-to-l 
difference due to expansion introduced by curvature of the reproducing 
cathode-ray tube voltage-to-luminance transfer characteristic. In 
the studio, the transfer characteristic relating scene luminance to 
output voltage for the image orthicon tube is similar in shape to the 
familiar H&D curve for film but is influenced by the level of illumina- 
tion incident on the image orthicon photo-cathode, by the ratio be- 
tween high-light and average scene luminance and by adjustment of 
image orthicon target voltage. It is possible to have compression of 
one range of luminance values and expansion of another all in the 
same scene. This problem again calls for careful control of over-all 
scene luminance distribution and for careful exposure and camera 

These troubles are caused in part by the electron redistribution 
process inherent in present-day operation of the image orthicon pickup 
tube. Several spurious effects which arise in the tube or associated 
equipment are even more objectionable at times because of their 
distinctive appearance. These effects of interaction between adja- 
cent areas include : 

Halo: A black area surrounding a bright high-light, resulting from a 
rain of low-velocity electrons emitted from the high-light area of the 
image orthicon target and particularly severe on highly polished 
jewlery, white clothing, and bald heads (see Fig. 1). 

Image orthicon ghost: A spurious, displaced image of a high-light 
area, most noticeable with severe contrast between high-light and 
background, resulting from high-velocity secondary electrons emitted 
from the high-light area on the image orthicon target (see Fig. 2). 

Clouding: An electronic fogging or mottling of large dark areas, 
similar in effect to lens flare, particularly severe where excessive con- 
trast exists between large dark and large light areas and aggravated 




Fig. 3-A. Clouding. 

Fig. 3-B. Clouding. 


Fig. 3. Clouding 

The inability of the image orthicon to maintain high contrasts over large areas 
is indicated by the clouding within the silhouette in A. The halo effect serves to 
maintain a dense black over small areas but beyond its reach, the larger areas ar<> 
a milky gray, spotted with dynode spots. Note that, although the halo effect 
holds small area contrasts, it is in the nature of all or nothing, so in other than a 
silhouette effect, severe distortion in luminance transfer would result. Reduc- 
tion of the contrast from 35 to 1 to only 5 to 1, shown in B, enables a more uni- 
form black reproduction. This clouding effect will be recognized as being similar 
to lens flare although it is purely electronic in this illustration. 

by the tendency of multiplier electrode spots to show through in dark 
areas (see Fig. 3). 

Streaking: A dark or light horizontal streak across the picture in 
line with excessively bright high-lights or long, heavy scenery lines, 
usually resulting from improper low-frequency characteristics in the 
transmission system (see Fig. 4). 

These effects can be reduced or adequately hidden by the same care- 
ful control of staging, lighting and camera operation called for in 
working within the usable total contrast range and in obtaining good 
luminance transfer characteristics. 

In the matter of detail resolving capability, television falls between 
35-mm and 16-mm motion pictures. Good resolution is aided by the 
same careful staging and lighting called for above, as the camera can 
then be adjusted to peak performance rather than to a compromise 
which would accommodate a range of conditions. For example, some 
of the effects listed under clouding may be hidden by beam def ocusing ; 
the image orthicon ghost, by image def ocusing; control of staging 
arid lighting makes it unnecessary to throw away resolution to hide 
such effects. 


As the various spurious effects and the distortion in luminance trans- 
fer characteristics are greatly aggravated by overexposure, principal 
objectives of studio practices must be the maintenance of uniform 
luminance ranges and scene content throughout a production and the 
establishment of conditions which are within the capabilities of the 
television facilities. From the transmission viewpoint only, a very 
"flat" scene is the easiest to handle. From the equally important 
artistic standpoint, however, as much contrast as possible is desired 
to provide scope for artistic expression, simulation of depth and 
establishment of mood. The technical group must sometimes relax 
technical quality requirements where special effects are desired mo- 
mentarily ; the staging and direction groups, on the other hand, must 




Fig. 4- A. Streaking. 

Fig. 4-B. Streaking. 


Fig. 4- Streaking 

The white stair risers when aligned so as to parallel the television scanning lines 
produce pulse signals having large energy content on a number of scanning lines. 
In a video amplifier having improper low-frequency phase or amplitude response 
characteristics, a transient results, producing a streak across other portions of the 
picture. The streak may be of either the same or opposite polarity depending on 
whether leading or lagging low-frequency phase response is involved, for example. 
In the studio, the contrasts of large horizontal elements may be held down or the 
length of such elements along the scanning lines may be minimized by turning 
the element at an angle as in B. 

become accustomed to somewhat more restrictive conditions than are 
prevalent in motion picture or legitimate stage production. Be- 
tween the technical and artistic viewpoints lies an area for funda- 
mental compromise, the boundaries of which working rules must seek 
to establish. 

In practical production, the further problem of time and space 
separation among the various activities must be considered. The 
scenery may be prepared and costumes selected well ahead of the per- 
formance date. Lighting is planned in advance but adjusted after the 
prefabricated scenery with accompanying properties is set up in the stu- 
dio. Camera facilities often are activated only during late stages of stu- 
dio rehearsal. Thus, it is necessary that each major production oper- 
ation be guided by rules which allow its performance independently 
but with assurance that the work will fit together as a co-ordinated 
whole. In the following sections, rules for each production phase are 
grouped together for convenience in reference with a discussion of the 
factors involved following each of the groups. Under the subject of 
staging, which includes scenery preparation, selection of properties, 
costuming and make-up, the reflectance of materials is controlled. 
Lighting practices are controlled as to illumination levels and distri- 
bution, this being the planning phase of lighting as compared to the 
mechanical adjustment which is often considered a staging operation. 
Methods of camera operation are guided to take advantage of con- 
trolled scene conditions in obtaining the best possible camera per- 


1. The reflectance of large set areas and objects should be held 
between 25% and 50% for high-key scenes; between 10% and 25% for 
low-key scenes. Small details may range from 3j^% (maximum 
black) to 70% (maximum white) reflectance. A 10-step, 20:1 gray- 
scale step wedge should be used as a reference standard. 

2. Where colors are used, a limited number of samples should be 








50% -*- r i 







i i 

1 2 i 

201 5 



g! ! 
s i 1 i 



i 5 

101 r -i 

IO *-_L_I - 












TEP 10 









Fi0. 5. Reflectance Values: Recommended reflectance value ranges for large 
and small scenery areas or objects are shown with respect to a 10-step, 20-to-l, 
reference gray-scale step-wedge. The total range for large scenery areas is held to 
5 to 1 (approximately 10% to 50%) allowing for some additional variation from 
illumination distribution. Reflectance values indicated here are to be measured 
through a camera system by comparison with the calibrated gray-scale step- 
wedge, thus taking into account spectral distribution and surface texture. 

selected and calibrated against the reference gray-scale by observation 
on a picture monitor. 

3. Large adjacent areas should not differ in reflectance by more than 
2 to 1 (2 steps on the reference gray-scale) . 

4. Large monotone areas, particularly dark ones, should be avoided 
or broken up with simple patterns, tree leaves, shadows, etc. 

5. At least a small area of both extreme white and extreme black 
should be included in every scene to aid camera level adjustment. 

6. Very dark or very white long horizontal lines or structures 
should be avoided. 

7. Highly reflective areas or objects should be avoided or placed in 




diffuse light against a light background and, if necessary, treated to 
reduce reflectance. 

Discussion of Staging Practice 

The reflectance* of scenery elements is keyed to performer's skin 
reflectance which is on the order of 30% to 40% 13 (Fig. 5). If large 
background areas are too light, electron redistribution within the 
image orthicon camera tube from such areas may cause faces to be 
darkened ; if the scenery is too dark, electron redistribution and spu- 
rious effects in the camera tube arising from the faces may disturb the 
background. The total range of reflectance for large areas covers 
only a 5-to-l ratio (10% to 50%), allowing for some additional varia- 
tion in over-all luminance to be contributed by adjustment of illumi- 
nation distribution. 

The use of color in staging is unnecessary for monochrome television 
transmission. However, if used carefully as merely another way of 
obtaining a certain gray tone on the picture tube there are no ob- 
jections to its use. Prior to the establishment of a color-sample pre- 
calibration procedure, there was at least one case where a studio set- 
ting was carefully worked out in beautifully contrasted colors only 
to appear on the air as a complete monotone (the designer's red face 
produced the only different tone). The human eye tends to confuse 
color contrast with brightness contrast and, unaided, is often a mis- 
leading judge of staging propriety. One technique found to be safe 
is to paint scenery in graded tones of a single pre-calibrated color. 
Color is in television scenery to stay, it being felt by many that a more 
natural atmosphere is provided for the performers. Although this is a 
moot point, the practical approach is to use color but to control it in 
terms of easily determined pre-calibrated gray response. 

Electronic spurious effects in the camera tube are most noticeable 
where large contrasts and large dark areas are present. Bright ob- 
jects in front of dark backgrounds are a severe problem. The re- 
flectance of starched white cloth may run as high as 90%; that of 
jewelry, musical instruments and other polished objects may run 
close to 100%, exhibiting specular reflectance. These values are 
almost three times the reflectance of skin. In many cases, the high- 
light area itself may be controlled by: painting-off kitchen appli- 
ances; using blue, tan or off-white clothing and stage papers; using 

* Reflectance is taken here to indicate nonspecular reflectance averaged over 
the range of wavelengths accepted by the type 5820 image orthicon tube, and 
measured by comparison with the reference gray-scale, observed through a cam- 
era channel. 




Fig. 6-A. Excess top-light. 

Fig. 6-B. Excess top-light. 


Fig. 6. Excess Top-Light 

Illustration A was photographed from a camera control monitor; B was photo- 
graphed directly on the studio floor with the same lighting. The increase in con- 
trast through the television system is noted in comparing the two photographs. 
The shadow detail present in the direct photograph is lost in transmission through 
the system. Lighting contrasts which can be used for direct photography are 
excessive in television. Strong overhead lighting is particularly to be avoided. 

matte surface stage photographs; or coating metallic surfaces with 
spray-wax or talcum powder. Bald heads may be toned down 
with make-up. In other cases, the background surrounding an un- 
controllable high-light such as a piece of jewelry or a candle flame may 
be kept very light, being graduated off into darker areas of the scene. 
It must be realized, of course, that maintenance of a limited range 
of reflectance among portions of a stage setting is only part of the 
story. The light which actually reaches the camera is a function not 
only of scenery reflectance but also of illumination level, which in 
turn must be held within its own set of limits. 


1. The first step in lighting a television scene should be to establish 
a diffuse and uniform base-light throughout the working area, includ- 
ing backgrounds. Measured with a photocell meter aimed toward 
all camera positions from the various performer positions, the follow- 
ing levels should be obtained for use with the type 5820 image orthi- 
con equipped with a Wratten No. 3 filter : 

With 4500 K (degrees Kelvin) white fluorescent light, 

100 101m/sqft* 
With incandescent light (2870 K), 120 10 Im/sq ft. 

Components measured with the meter aimed vertically should not 
exceed the components measured with the meter aimed horizontally. 

2. To provide depth, to separate objects and to add artistic interest, 
several types of effects-light should be added, usually from directional 
incandescent fixtures. The level of such components may be set to 
approximately the correct values by measurement with a photocell 
meter pointed toward the fixture from the performer positions with all 
other lights turned off, but should be set finally on the basis of appear- 
ance on a picture monitor. 

Back-light: Should be directed from the lowest possible rear angle 
with an intensity between 1 and 1J^ times base-light level. Vertical, 
top-light is not back-light and should be avoided. 

* Lumens per square foot is numerically equivalent to the older term, foot- 




Fig. 7-A. Fluorescent base-light (100 ft-c including eye-light, E below). 

Fig. 7-B. Incandescent back-light (1 X base). 


Modeling-light: Should be directed from a side-front position; may 
be adjusted to just cause shadows and will usually require an intensity 
of between 3/2 an d 1 times base-light. 

Key-light: Similar to modeling but slightly more prominent to 
give effect of a predominate source; may be from 1 to 2 times base- 
light level with base-light from the direction of the key-light fixtures 
reduced accordingly. 

Eye-light: A very small amount of light to provide sparkle in a per- 
former's eyes and to supplement the base-light for close-ups; usually 
from % to 1 times base level. 

3. Where special lighting effects such as spot-lighting, lights-out 
dramatic sequences, or moonlight effects are required, base-light may 
be lowered to a minimum of Y normal level and any special effects- 
lights then adjusted to bring total illumination level at the point of 
interest between 1 and \y% times the normal base-light level (meter 
aimed at camera). 

4. Fluorescent (4500 K, white) and incandescent light sources 
may be intermixed without impairing color-response where cameras 
are equipped with Type 5820 or 5826 image orthicons and blue-cutting 
filters such as Wratten No. 3. Incandescent sources may be dimmed 
to J4 normal meter reading without impairing spectral response. 

5. Self-luminous objects or areas such as exposed lamps, rear- 
lighted windows, or rear-projection high-lights shall be held (by dim- 
ming or other means) to produce, on a meter held a few inches from the 
area, a reading which does not exceed % to 1 times the normal base- 
light level. 

Discussion of Lighting Practice 

Base-light corresponds to what often is called fill-light in motion 
picture practice. However, because of the contrast range and spuri- 
ous response characteristics of the television system, this is the most 
important of all lighting components and it must be given the prom- 
inence of first attention ahead of key- or other effects-light. The 
term, base-light, connotes the basic importance in television lighting 
of covering all picture areas with a very uniform illumination level. 
It is especially important that this base-light be uniform over the 
entire working stage area as viewed from all possible camera view- 
points and that it include low-angle front-light components. Where 
the reflectance of scenery and set fixtures is controlled, uniform, diffuse 
base-lighting will insure similar working conditions for all cameras on 
all sequences of a production. It has been found that these condi- 
tions can be adequately fulfilled with either fluorescent or incandescent 



Fig. 7-C. Fluorescent base-light and back-light. 

Fig. 7-D. Incandescent modeling-light alone (1.7 X base, strong enough to be 



sources of light so long as good diffusion and even distribution are 
obtained. Mixtures of lighting types for this function are permissible. 
Excess top-light is to be particularly avoided as it is a prime source of 
electron halos and ghosts as well as darkened eyes and faces (Fig. 6). 

Television pictures obtained with good base-lighting alone are tech- 
nically good, but artistically incomplete. To achieve artistic quality, 
a relatively small amount of effects-lighting is required as compared 
with movie, legitimate stage or photographic practice. The television 
system cannot tolerate violent lighting contrasts; but, on the other 
hand, will produce pleasing results with relatively small intensity 
differences. The important technique is to use differences in quality 
or direction rather than brightness differences alone, to obtain the 
effect. There are an infinite variety of effects-light possibilities, and 
it is practical to catalog only a few very common types, with recom- 
mended intensities to indicate general technique (Fig. 7- A G). In 
general, it is necessary to look at the result on a picture monitor 
not trusting direct visual observation on the set, to determine the 
effectiveness of any special lighting arrangements The eye accepts 
too wide a range of contrast to be a reliable measuring instrument. 
It is desirable that the lighting directors become accustomed to this 
practice, as one job they must often perform is to correct or supple- 
ment any staging conditions which do not register as desired. 

As to color response, camera-tube developments have simplified this 
problem greatly. 3 Also, the subjective tolerance for visible color- 
to-gray distortion is actually somewhat greater than has been gen- 
erally believed; most of television's notorious earlier troubles arose 
from erratic sensitivities of pickup tubes to nonvisible light com- 
ponents. At the present state of the art, it can be said that any light 
which appears reasonably white to the eye will produce satisfactory 
color-to-gray rendition on television. With a stage set having con- 
trolled reflectances, and with lighting which has been arranged to 
have a uniform base-light level with carefully adjusted artistic effects, 
the camera operating personnel are in a position, with a few additional 
precautions, to secure good results. 


1. The type 5820 image orthicon, used with a Wratten No. 3 filter, 
will require a normal lens aperture of from //8 to //16 when recom- 
mended light intensities are used. The exact lens setting should be 
adjusted to the point where the signals corresponding to maximum 
high-light areas just start to decrease in amplitude on the waveform 




Fig. 7-E. Fluorescent base-light, back-light and modeling-light. 

Fig. 7-F. Incandescent eye-light (including additional base-light measured as 

part of A). 


2. All cameras used on a particular production shall be carefully 
adjusted to give matched gray-scale rendition against actual set 

3. Image orthicon ghosts may be minimized by keeping the offend- 
ing high-light in the center of the screen or, where necessary, by elec- 
tron-image defocusing. 

4. All control room monitors should be equipped with picture tubes 
of similar phosphor color and of similar voltage-to-luminance transfer 
characteristic. They should be adjusted to just go black with a 
blanking signal set to reference black amplitude. If in doubt, camera 
balance comparisons should be made on a single (line-output) moni- 

5. Signal levels should be carefully monitored to maintain true 
black-and-white signals at their respective reference levels, using the 
picture monitor to judge which peaks may be considered spurious or 
to judge conditions where no full peak signals exist. 

Discussion of Camera Operation 

Although some of the present-day image orthicon tubes will produce 
very satisfactory results from sets illuminated with only 20 to 30 
Im/sq ft, there is about a 3:1 variation in sensitivity among tubes. 
To accommodate this range, to provide a margin for filter absorption 
and to accommodate special lens conditions, light intensities of the 
order of 100 Im/sq ft are recommended. Present practice is to use the 
lens stop to regulate exposure, contrary to the established motion 
picture practice of setting lens stop to achieve a required depth of 
focus. As pickup tubes become more uniform in sensitivity, this 
practice will undoubtedly be adopted in television. If necessary, 
base-light may be reduced to 50 Im/sq ft with appropriately wider 
lens apertures and if a uniform, diffuse distribution is maintained. 

Although the sensitivity varies, the color response is quite uniform 
among present-day type 5820 and 5826 image orthicon tubes. The 
response of these tubes to infrared components is negligible and it has 
been found that a simple filter to remove the monochromatic mercury- 
line radiation present in fluorescent light enables satisfactory color 
match among various cameras and under various types of light. The 
Wratten No. 3 (or No. 6) 10 filter has been found to be a good compro- 
mise between effective blue-rejection and loss in the transmission 
band, a filter factor of one lens stop being applicable. 

Color response differences now seem to be less of a factor in gray- 
scale matching among cameras than lens flare, scene luminance con- 
tent or camera misadjustment. With excess exposure, a considerable 

262 RICHARD S. O'BRIEN September 

Fig. 7-G. Full base-light, back-light, modeling and eye-light. 

variation in luminance transfer characteristic may result from adjust- 
ment of image orthicon target potential. The more positive the volt- 
age above beam cutoff, the greater the range of input luminance which 
can be accommodated. At the same time, electron redistribution 
effects decrease as the target voltage is made more positive so there is 
much less change in over-all luminance transfer characteristic as the 
average scene luminance is varied. A limit to the potential of the 
target above cutoff is set by beam-current noise ; beam current being 
increased in proportion to target voltage to discharge all high-light 
areas. However, the fact that camera target voltage adjustment 
does influence over-all transfer characteristic may be used within limits 
to correct an unfortunate staging or lighting circumstance. These 
effects, though more apparent in the close-spaced tubes such as the 
type 5826, have been found to apply in a limited way to the type 
5820 which has a wider target to mesh spacing. Video gain and 
blanking level adjustments also affect over-all luminance transfer 

Television signal monitoring techniques are similar to audio prac- 
tice in that two devices are used : one to measure levels, the other to 
determine quality. However, the technique of monitoring by obser- 
vation of waveform and picture monitors, lacks the years of refine- 


ment and widespread operational use which lie behind use of the audio 
volume indicator meter and the loudspeaker. At the present, tele- 
vision monitoring techniques are somewhat primitive. To make the 
best use of the present facilities in balancing gray-scale, average signal 
level, and black-and-white peak level among cameras, picture moni- 
tors must be carefully set up and waveform monitors carefully cali- 
brated and interpreted. Where definite maximum whites and blacks 
exist in a scene, level monitoring is straight-forward. Where glints 
or small insignificant blacks exist, they may be adjusted to exceed 
their respective reference levels requiring judgment of their im- 
portance as viewed on the picture monitor. Where a very flat image 
is desired there may be no peak levels of reference amplitude at all; 
the video gain and blanking controls again have to be set by judgment 
of the image seen on the picture monitor. 

In such cases, the faces of performers should be used as the basis of 
judgment, the voltage waveform corresponding to such areas normally 
being held between J^ and % f reference white level. It is unfor- 
tunate that no television equivalent of the standard audio volume 
indicator meter is available as yet. The importance of careful and 
accurate level monitoring is realized when it is considered that current 
practice calls for setting black peaks to a level which is 10% of total 
picture amplitude, a value which is only little more than the limiting 
accuracy in reading an oscilloscope. The dependence of image quality 
on careful camera adjustment and operation in no way lessens the 
requirement for careful staging and lighting practice. The technical 
job is made much simpler if the preceding steps have been done so as 
to minimize some of the problems, allowing concentration on the 
remaining ones. Achievement of uniform staging and lighting condi- 
tions makes possible optimum camera adjustment to obtain good 
results from all cameras throughout a show. 


An understanding of the technical characteristics of the present 
monochrome television facilities makes possible control of staging, 
lighting and camera operation to achieve technically correct and ar- 
tistically pleasing images. In particular, characteristics of the tele- 
vision pickup and transmission facilities establish boundaries on over- 
all range of luminance values, on luminance differences among adja- 
cent areas, and on the shape and arrangement of scenery features. By 
setting forth working rules grouped for the various production activ- 
ities which are often separated in both time and location, production 
personnel are enabled to work separately toward a unified final result. 


It is realized that working rules cannot cover all cases that they are 
subject to early obsolescence in such a rapidly changing art as tele- 
vision. However, by pointing out the general approaches, such 
rules enable the experimentally minded production team to guide their 
very commendable striving for better effects away from known blind 
alleys or even to make use of the limitations themselves. 

Results which exploit the capabilities of the present television facili- 
ties to the greatest possible extent can be obtained in everyday opera- 
tion when the system characteristics are known and respected. 


The interested, helpful supervision of this work by William B. Lodge, Vice- 
President in Charge of General Engineering for CBS, and by Howard A. Chinn, 
Chief Audio- Video Engineer for CBS, is gratefully acknowledged. The excel- 
lent co-operation given hi the course of the investigations by the various opera- 
tions departments of WCBS-TV is very much appreciated. The author is 
particularly indebted to Ted Lawrence, Quality-Control Supervisor for CBS-TV, 
for the numerous contributions made in working out and actually applying these 


1. O. H. Schade, "Electro-optical characteristics of television systems," RCA 

Rev., vol. 9, nos. 1, 2, 3, 4; Mar., June, Sept. and Dec. 1948. 

2. R. B. Janes, R. E. Johnson and R. S. Moore, "Development and perform- 

ance of television camera tubes," RCA Rev., vol. 10, no. 2, June 1949. 

3. R. B. Janes, R. E. Johnson and R. R. Handel, "A new image orthicon," RCA 

Rev., vol. 10, no. 4, Dec. 1949. 

4. H. Kozanowski, "How to get the best picture out of your image orthicon 

camera," Broadcast News, RCA Publication, no. 54, Apr. 1949. 

5. Tips on Use of Image Orthicons, (booklet), published by RCA Tube Dept., 


6. R. M. Fraser, "Motion picture photography of television images," RCA Rev., 

vol. 9, no. 2, pp. 202-217, June 1948. 

7. H. E. Kallmann, "The graduation of television pictures," Proc. I.R.E., vol. 

28, no. 4, pp. 170-174, Apr. 1940. 

8. D. G. Fink, "Brightness distortion in television," Proc. I.R.E., vol. 29, no. 6, 

pp. 310-321, June 1941. 

9. A. H. Brolly, "Television studio lighting," Jour. SMPE, vol. 53, pp. 611-622, 

Dec. 1949. 

10. Richard Blount, "Lighting distortion in television," Jour. SMPE, vol. 53, pp. 

625-634, Dec. 1949. 

11. H. M. Gurin, "Illumination for television studios," Tele-Tech, Pt. I, vol. 8, 

no. 9, pp. 54-73, Sept. 1949; Pt. II, vol. 8, no. 10, pp. 34-57, Oct. 1949. 

12. R. L. Kuehn, "Television photometry and optical background," Tele-Tech, 

vol. 8, no. 7, pp. 24-50, July 1949. 

13. R. M. Evans, An Introduction to Color, 340 pages, John Wiley, New York, 

N.Y., 1948. 

Motion Picture Instruction 
In Colleges and Universities 

A Follow-up Study of the 1946 Report by John G. Frayne 



4s THE HOT WAR of the 40's ended, there was much to be said about the 
_/JL growing use of film in education and about the teaching of film produc- 
tion in colleges and universities, with comprehensive curriculums supple- 
mented by modern and suitably designed equipment. These views, in fact, 
were reflected in the JOURNAL article by Frayne cited in the note below. As 
Chairman of the SMPE Motion Picture Instruction Committee, Frayne 
reported the motion picture courses taught in the 102 institutions, colleges 
and universities which answered his questionnaire. The purpose of the 
study was to report to the Society what courses were being taught and 
where. The article closed with comments as to the possible value of this 
education to motion pictures as a profession and noted in particular the lack 
of technical courses. As of three years later three years of unparalleled 
expansion in higher education the present investigation proposes to follow 
up the Frayne study for the purpose of indicating possible trends and offer- 
ing a reasonably definite idea of the present state of instruction in motion 
pictures in American colleges and universities after this period of relatively 
lush growth. 

In order to compare the two reports simply, the Frayne breakdown of 
courses is followed as originally conceived. The Frayne report (1946) ap- 
pears on the left side of the tabulation immediately following, and on the 
right is the follow-up study (1949). In addition to the information about 
units and semester hours included in the 1946 report, the 1949 report has a 
brief description of each course. 

Because certain schools not included in the 1946 study have since then 
introduced courses, and because schools which have increased the number of 
their courses since that time would not appear in the parallel type of report- 
ing, two additional summaries are necessary for a more comprehensive con- 
sideration of the growth in the period. These listings below, are shown 
on p. 273 through p. 277. One is a recapitulation of the courses dropped since 
1946, and the other is a listing of schools which were not reported in the 
Frayne study and which have introduced motion picture courses since 1946. 
Due to the newness of the field, it is impossible to say how many institutions 
offering motion picture courses are still not included in this report. 

A CONTRIBUTION: Submitted December 15, 1949. This follow-up study of 
"Report of the Committee on Motion Picture Instruction" by John G. Frayne, 
Jour. SMPE, vol. 47, pp. 95-106, Aug. 1946, was instigated by the American 
Educational Theater Association's Committee on Film, Radio and Television 
under the Chairmanship of Kenneth Macgowan and was reported at that As- 
sociation's 1949 National Convention. 





8 f 4 M 8 1 

s * 5 a 'S 

OT2 -"8 * 

-1 89 H A A 

e -rt 8 H o 

Sill fa 

g 3 2 


P.S vs * 




-o ^ 

s.^^a'e g 



'a-a -s s -^ 2 

P C n3 'C 

S g:i a S 
a |f|- 



O> +3 


.5 o 


e.2'-S ft *3 
II ,. 
















fl w 2 

Igjjj , 

8 'i 8 a s * 


g 1 ^ 

1 I a- 1 I 

2| M Hi^ 

Iff!!.? * 

Illl^l s 

CJ ^ O C ^ C3 03 


.fa . j/ S .2 1 - 

rr. 02 r$ 






'-5 C 

3 ' 

- 1 s 


^ I 



3 1 2 "i 

^^Sl-s * 
N ^| t s H .| 5 

a r J:Jil - 

S - M' S 

I'll Hi 


S I 

1 ! 







3^ -*<* . 

*3 S.. a> 


G a .s 


3? ft? 




' "2 



" w 




S 3 









i s 









g "c 













? " 
S .. 






ci ^ 


^ b^ 




C ft! 





Covered i 



Z o 



6 ^ 


H I 


Courses Dropped by Schools Since the 1946 Report 

School Course 


New York University Motion Pictures 1 & 2 

Ohio State University Cinematography 

University of Denver Motion Picture Making 


Baylor University Photography 

Drake University News Photography 

New York University Motion Pictures 3 & 4 

St. Olaf College Photography and Art 

University of Oregon Rudiments of Photographic Journalism 

University of Southern California Cinema 90 & 91 

Sound Recording 

New York University Motion Pictures 9 & 10 

Oregon State College Education 533 

Motion Picture Distribution 

New York University Motion Pictures 19 & 20 

Pennsylvania State College . . . Motion Picture Distribution 

Economic Problems in Motion Picture Production and Exhibition 

University of Southern California Cinema 250AB 

Courses Reported Since the 1946 Report 

(Includes Schools Not Reported by Frayne) 


Baylor University Drama 388 The Film 

College of the City of New York Films 13 Fundamentals of Film Production 
Drake University Ed. 108 Audio- Visual Education Materials and 

New School for Social Research . Elements of Cinematography 

New York University Cinematography 

Ohio State University Motion Picture Photography 

Stanford University The Technique of the Motion Picture 

University of California at L. A. . Motion Picture Photography and Sound 
University of Denver Ed. 336 Survey of Audio- Visual Materials and 

Film Techniques 

University of Minnesota .... Motion Picture Photography 
University of North Carolina . . Motion Picture Production 
Western ReserveJUniversity . . Motion Picture Production 
West Virginia University .... Ed. 251 Cinematography 


Boston University ....... News and Feature Photography; 

Advanced News and Feature Photography; 

The Preparation of Photographic Materials for 

Visual Education 

Brooklyn College of the City of Design 45 Photography I; 
N. Y. Design 46 Photography II 

College of the City of New York Motion Picture Photography; 

Advanced Motion Picture Photography 

Columbia University Science 169P Photography for Teachers 

Cornell University An introductory course in photography 




Courses Reported Since the 1946 Report, cont'd 

Depauw University 
Drake University . 

Indiana University 

Kansas State College of Agr. 
App. Sc. 

Miami University 

Ohio State University .... 

South Dakota State College of 

Agr. & Mech. Arts 
Texas Christian University . . . 
Tulane University (Newcomb 


University of Mississippi .... 
University of New Mexico . . . 
University of North Carolina . . 
University of Oregon 

University of Southern California 
West Virginia University .... 

Sound Recording 

Brigham Young University . . . 

College of the City of New York. 

Indiana University 

Ohio State University 

Pasadena City College 

University of Southern California 

Motion Picture Film Editing 

Baylor University 

College of the City of New York . 

University of California at L. A. . 
University of Southern California 

Motion Picture Projection 
Boston University 

Brigham Young University . 

Iowa State College 

Oregon State College. . . . 
University of Virginia . . . 

West Virginia University . . . 

Motion Picture Distribution 
College of the City of New York 
Indiana University 

New School for Social Research 
University of Virginia .... 
West Virginia University . . . 

Composition and Photography 
Pictorial Journalism; 
Camera Journalism 
Creative Photography; 
Elementary News Photography; 
Practical Work in News Photography; 
News Photography; 
Picture Editing 


Engineering Photography; 
Scientific Photography 


A Camera Department 

Ed. 19d/e The Use of Photography in Teaching 

Art 87 & 88 Photography 

Elementary Photography 

Physics 161 Rudiments of Photography; 

Journalism 451, 452, 453 Graphic Journalism 

Fundamentals of Photography 

Physics 216 Photography 

Radio Production and Recording; 

Radio Sound Recording 

Film Music and Recording 

Radio Broadcasting 278AB 

511, 512 (sound-on-film recording in 16-mm) 

Radio Controls Laboratory 40 

140B Sound II 

Film and Television Production 
Motion Picture Editing; 
Advanced Motion Picture Editing 
Motion Picture Editing 165AB 
135B Editing II; 
185AB Editing III, IV 

The Operation and Maintenance of Audio- Visual 

Education 275 Audio-Visual Instruction 

Audio- Visual Methods in Education 

Audio- Visual Teaching Aids 

Ed. 57 Visual and Auditory Materials of Instruc- 

Ed. 221 Audio-Visual Resources for Instruction 

Distribution and Publicity in Motion Pictures 
Administration of a College Center of Audio- 
Visual Materials 
Operating the Film Library 

Ed. 159 Administration of Audio- Visual Programs 
Ed. 322 Organized Programs of Audio- Visual In- 




Courses Reported Since the 1946 Report, cont'd 

Economic Problems in Motion Picture Production and Distribution 

Boston University Administration of a Motion Picture and Audio- 
Visual Aids Dept. 

Film Production Methods 

Cinema and Society; 

Unit Management; 

Studio Production Control 

New School for Social Research . 
University of Southern California 

Film Processing (Still) 

South Dakota State College of 

Agr. & Mech. Arts 
University of Colorado 

Film Processing (Motion Picture) 
University of Southern California 
Motion Picture Acting 
New School for Social Research . 
University of California at L. A. . 

Screen Writing 

Boston University 

College of the City of New York 

Columbia University 

New School for Social Research . 

New York University ..... 

Stanford University 

University of California at L. A. . 
University of Southern California 

Western Reserve University . . 
Motion Picture Directing 
College of the City of New York. 
New School for Social Research . 

New York University 

Stanford University ...... 

University of California at L. A. . 
University of Southern California 

Motion Picture Lighting 
University of Southern California 
Educational Film 

Boston University 

College of the City of New York 
Columbia University. . . . . . 

Indiana University 

Photography 58AB 
News Photography 

101AB Laboratory Practices and Procedures 

Acting for Film, Television and Radio 
Acting for the Motion Pictures 

Writing of Motion Pictures and Filmslides 
Motion Picture Writing; 
Advanced Motion Picture Writing 
F. A. Motion Picture; 
Scenario Writing and Production 
Basic Screenplay Writing; 
Advanced Documentary Writing; 
Feature Screenplay Writing Seminar 
Writing the Screen Treatment 
Technique of the Motion Picture 
Writing for the Screen 166AB 
Screenwriting I, II, III, IV; 
Educational Screenwriting; 
Documentary Screenwriting; 
Seminar in Screenwriting 
Practices in Script Writing 

Motion Picture Directing 

Film Direction 

Intermediate Motion Picture Production 

Technique of the Motion Picture 

Fundamentals of Motion Picture Direction 

Cinema Directing I, II, III, IV; 

Seminar in Motion Picture Direction 

115AB & 165AB (Camera Classes) 

The Use of Audio-Visual Aids in Education 
The Documentary Film as an Educational Tool 
Science Films; 

Production of Educational Motion Pictures 
Utilization of Audio-Visual Materials; 
Selection of Audio-Visual Materials; 
Administration of Audio- Visual Materials; 
Production of Audio- Visual Materials; 
Administration of a College Center of Audib- 

Visual Materials; 
Seminar in Audio- Visual Materials; 

276 JACK MORRISON September 

Courses Reported Since the 1946 Report, confd 

Indiana University, cont'd. . . . Research in Audio- Visual Materials; 

Thesis in Audio- Visual Materials; 
Workshop in Administration of the Audio- Visual 

Aids Program 

Louisiana Polytechnic Institute . Use of Audio-Visual in the Classroom 
New School for Social Research . Audio- Visual Aids in Education 

Oregon State College Construction and Use of Audio- Visual Aids 

University of California at L. A. . Educational and Documentary Film Techniques 

University of Denver Survey of Audio-Visual Materials and Equipment; 

Survey of Instructional Motion Pictures; 
Administration and the Supervision of the Audio- 
Visual Program 
University of Kentucky .... Visual Teaching; 

Motion Pictures in Education 

University of Southern California 276AB Workshop in Educational Film Production 
University of Wisconsin .... Methods of Audio- Visual Instruction 

Documentary Film 

College of the City of New York Fundamentals of Film Production 
University of California at L. A. . Nature and History of the Documentary Film 
University of Southern California 208 AB Documentary Production; 

Documentary Direction 


New School of Social Research . Graphics and Animation 

University of California at L. A. . Fundamentals of Motion Picture Animation; 

Animation for Educational and Documentary 

Animation for Entertainment Film 
University of Southern California 148AB Principles and Mechanics of Animation 

Film History and/or Aesthetics 

Boston University The History of Motion Pictures 

College of the City of New York The History of Motion Pictures 
New School for Social Research . March of Film; 

Seminar in Film Techniques 

New York University Introduction to Motion Pictures 

Purdue University English 52 The Art of the Motion Pictures 

Stanford University History and Aesthetic Development of Motion 


University of Connecticut . . . The Art of Motion Pictures 
University of Southern California Introduction and Survey of Motion Pictures 60AB ; 

Filmic Expression 105AB; 

Cinema History and Criticism 200AB; 

Seminar in Creative Cinema 274AB 
Wayne University History and Appreciation of Motion Pictures 

Film Appreciation 

Baylor University Introduction to Drama, Television and Film 

Columbia University Ed. 162 PRD, Photography and Radio Drama as 

Communication Arts; 

Ed. 209 MF, International Film Series 

Fordham University Motion Picture Appreciation and Criticism 

Miami University Motion Picture Appreciation 

New School for Social Research . Basic Principles of the Mass Communication Arts; 

March of Film 

New York University Motion Picture Literature 

Syracuse University Cinema Appreciation 

University of California at L. A. . Visual Analysis 

University of Delaware Theater, Film and Radio 

University of Denver Film Arts 


Courses Reported Since the 19J+6 Report, concUd 

University of Iowa Cinematography 

University of Kansas The Motion Picture 

University of Minnesota .... Film and Drama; 

Humanities 52 

University of Oregon Appreciation of Drama 

University of Toledo Appreciation of the Motion Picture 

Film Design 

New School for Social Research . Applied Stagecraft for Film and Television 

Otis Art Institute Motion Picture and Television Art Institute 

University of California at L. A. . Motion Picture Costume Design; 

Motion Picture Design and Draftsmanship 167AB 
University of Southern California Art Directing I, II, III, IV; 

Art Direction 210AB 

Survey of Film Production and Techniques 

Boston University Workshop in Motion Pictures and Visual Aids; 

Motion Picture and Television Film Production 

Columbia University Production of Educational Motion Pictures 

New School for Social Research . Film Production Methods 

Pasadena City College Stage Technology 

University of California at L. A. . Motion Picture Survey; 

Film Technique 

University of Denver Motion Picture Production 

University of Southern California Motion Picture Production Techniques 175AB 
University of Wisconsin .... Local Production of Audio- Visual Materials 


Baylor University Television and Film Workshop 207 ABC 

Boston University Research in Motion Pictures and Audio-Visual 


Audio- Visual Aids in Health and Physical Educa- 

Visual Presentation of Ideas; 

Principles of Motion Pictures and Audio-Visual 

Aids in Public Relations and Business 
College of the City of New York The Documentary Film in Labor Relations; 

Practice in Film Production 
Columbia University Audio- Visual Materials and Methods of Use; 

Lab Course in Audio- Visual Instruction; 

Administering the Use of Audio- Visual Materials 

New York University Advanced Individual Study 

Stanford University Stage and Screen 

Texas Christian University . . . Research Problems in Speech-Drama 
University of California at L. A. . Motion Picture Makeup; 

Elementary Motion Picture Workshop; 

Advanced Motion Picture Workshop; 

Summer Motion Picture Workshop 179CDE; 

Theory of Educational Film 
University of Southern California Motion Picture Technology; 

Cinematic Effects 153AB; 

Makeup for Motion Pictures; 

Public Relations in Motion Pictures; 

Unit Management; 

Production 205AB; 

Seminar in Motion Picture Engineering 211AB; 

Studio Production Control 225; 

Films for Television 

Wayne University Advanced History and Appreciation of Theater 



In conclusion, it may be safely said that despite the fast-growing 
attention given to the motion picture in education, the schools in- 
cluded in the Frayne study indicate little if any significant change in 
the teaching of the production of motion pictures in colleges and 
universities in the years from 1946 to 1949. Although the follow-up 
study reported 300 motion picture and "related" courses compared 
with Frayne's report of 86, it is possible that the number of insti- 
tutions offering a comprehensive major in motion picture production 
is less than a half-dozen. This is probably a reflection of expense 
of equipment, lack of personnel and antipathy toward "trade school" 

Frayne's point that few technological courses are offered continues 
to be well founded. While it may be argued whether motion pic- 
tures is an art, a profession or a craft, there still does not appear to 
be enough study of the technical phases to assure motion pictures 
becoming understood and used as the distinctive educational tool 
it promises to be. 

A great number of the courses added have been in audio-visual 
aids. It is obvious that the relationship between motion pictures in 
particular and audio-visual aids in general needs clarification. For 
example, one development to be noted in the follow-up study is the 
recurrence of a one-semester omnibus course which not only teaches 
a student all the phases of making a motion picture, but requires 
him to complete one as a project in the course. While this may serve 
a certain situation well, its effect on the development of motion pic- 
ture production should be scrutinized carefully. 

Finally, attention should be called to the lack of a generally ac- 
cepted nomenclature in the field. This lack seriously impaired the 
effectiveness of this as well as the Frayne study. Attention to 
nomenclature, course description in relation to curricular concepts, 
and clarification of relationships to visual aids would be well worth 
the while of the American Educational Theatre Association's Com- 
mittee on Film, Radio and Television and the University Film Pro- 
ducer's Association. 

Synchronous Recording 
On '/4-In. Magnetic Tape 



SUMMARY: This article discusses the problem of synchronizing motion pic- 
ture film with a sound track on standard %-in. magnetic recording tape. 
The equipment for synchronizing the tape with film is the major subject dis- 

THE USE of magnetic tape for recording motion picture sound 
tracks has by now aroused great interest within the film industry. 
The system of recording a sound track directly on optical film is 
unnecessarily costly and risky and will soon be obsolete. Only too 
frequently retakes are necessary because of failure on the part of a 
performer or in later film processing. Failure to get a perfect track 
results in a great waste of film, time and developing cost. Magnetic 
recording can replace film recording entirely for sound track work and 
will save the industry a great deal of money. 

Early work with magnetically recorded sound tracks was done with 
standard 35-mm motion picture film coated on one side. Later, the 
film was split down the center to save one-half of its cost. However, 
the cost of split 35-mm magnetic recording film is ten times as high as 
standard J^-in. unperf orated tape. This difference in cost makes the 
latter recording medium appear to be most desirable if it can be used. 
Not only can it be used, but it has several other important advantages 
over the perforated tape, aside from that of cost. Storage space is 
reduced by 7J^:1 over the split 35-mm magnetic tape. Recording 
time per reel is increased by 2^:1. Weight per reel is reduced by 

As everyone in the film industry knows, the problem of sprocket 
perforation flutter is a major problem which required considerable work 
to overcome. The use of J^-in. magnetic tape for sound track record- 
ing eliminates this problem as well. The manufacturers of magnetic 
recording materials have stated that the magnetic coating cannot be 
applied to 35-mm film as uniformly as it can on J^-in. tape. The 
greater uniformity obtainable on tape results in lower amplitude 
modulation of the recording and better high-frequency response. The 
35-mm film base has approximately nine times the stiffness of K- m - 

PRESENTED: April 24, 1950, at the SMPTE Convention in Chicago. 





tape. This greater stiffness results in poor head contact and conse- 
quent further amplitude and frequency response variations not found 
in using standard tape. 

To be of any use to the motion picture industry, it is necessary to 
synchronize the sound track with the picture. Since no sprockets can 
be used with J^-in. magnetic tape, one must magnetically or optically 
mark the tape so that it can be reproduced later at a controlled rate. 
The optical methods of tape synchronizing normally utilize bars or 
spots on the back surface of the tape as guides to control the speed 
of playback and hence allow it to be synchronized with the film. The 
systems thus far tried using this type of tape marking are satisfactory 
except during starting. Due to the fact that the photoelectric sens- 
ing devices do not know whether the tape at start is running faster or 










Fig. 1. Tape synchronizing system. 

slower than synchronous, it is impossible for the sensing system to con- 
trol the tape drive during starting. The lip-synchronous equipment 
described herein utilizes a magnetic marking system which has the 
advantage that it will without attention from an operator come to 
film-synchronous speed from standstill in a time corresponding to 
normal starting time for the associated recording equipment. 

Figure 1 is a block diagram of the tape synchronizing system devel- 
oped to be used with a standard Ampex Model 300 tape recorder. 
On the left side of Fig. 1 the block marked "Recorder" is a standard 
tape recorder without any modifications whatsoever. The block 
marked Fig. 2 is a small, lightweight, auxiliary unit which marks the 
tape magnetically during the time the original sound track is re- 
corded. It will be noted that the audio signal to be recorded is fed 
into this unit marked Fig. 2 and before entering the recorder has added 




to it an 18-kc carrier which is modulated by the frequency of the 
power line used to drive the picture recording camera. In Fig. 1 this 
is the input marked "60-cycle line input." However, the system 
will operate properly if some frequency other than 60 cycles is used. 
After making a recording with this setup, the tape contains the 
intelligence as well as 60 cycles riding on an 18-kc carrier. The rela- 
tionship between the 60-cycle signal and the audio intelligence is the 
same relationship as the picture on the film recorded by the film 
camera bears to the same 60-cycle frequency. Hence, if during play- 
back the 60-cycle signal on the magnetic tape can be held in fixed 
relationship with the power-line frequency driving the projection 






Fig. 2. Modulated 18-kc oscillator and mixer. 

camera, the sound and picture will remain in a synchronous relation- 
ship. Figure 2 shows the electrical circuit of the modulator unit used 
with the recorder. It consists of a push-pull 18-kc oscillator, which is 
amplitude modulated by the 60-cycle line frequency through the 
modulation transformer, TV Mixing of this modulated 18-kc carrier 
with the incoming audio frequency is accomplished by the balanced 
attenuator network shown in the upper portion of the figure. 

Referring to Fig. 1 again, the block marked "Playback" represents 
the recorder when used as a playback machine. When a tape, which 
has been previously recorded with the equipment shown on the left 
side of Fig. 1, is played back, the output from the machine contains 
the audio intelligence and the 18-kc amplitude modulated carrier. 




This signal is fed into the equipment shown in Fig. 3 which is the 
differential speed detector and power amplifier. This equipment 
controls the speed of the playback so that the rate at which the tape 
travels is exactly the same as the film playback. In Fig. 3 the block 
breakdown of the equipment used for this purpose is shown. In the 
upper left corner, the output of the playback amplifier enters the speed 
control system. In passing through the 18-kc rejection filter, the 
carrier and its modulation are removed, leaving only the audio intelli- 
gence at the output terminals. Before the 18-kc rejection filter, a tap 
is made which takes the 18-kc modulated carrier and the audio to the 






0^ * 

50 WATT 

60 t- * II5V SO WATTS 










Fig. 3. Differential speed detector and power amplifier. 

input of the 18-kc % selective-limiter amplifier. Output of this ampli- 
fier feeds a conventional full-wave detector circuit which delivers 
at its output the approximately 60-cycle line frequency. The word 
"approximately" is used in this case because thus far we have not 
considered where the correction in the tape speed occurs. The 60- 
cycle output from the detector drives the power amplifier which 
delivers approximately 10 watts of power at 60 cycles. This ampli- 
fier has in its output stage two 6V6-type tubes. The heart of this 
synchronizing equipment lies in the use of the two synchronous motors 
shown at the output of the 60-cycle, 10-watt amplifier. The syn- 
chronous motor which receives its input from the 10-watt amplifier 


is mounted rigidly. The output shaft of this motor is directly con- 
nected to the output shaft of an identical motor, but the second motor 
is mounted in such a way that its stator is free to rotate through ap- 
proximately 180 degrees. When its stator rotates it operates the 
potentiometer shown to the right of these two motors. The second 
motor obtains its input from the 60-cycle supply used in operating the 
projector which is projecting the film associated with the sound on the 
tape being reproduced. The potentiometer driven by the stator of 
the second synchronous motor supplies a variable d-c voltage to the 
variable-frequency oscillator shown in the lower left corner of Fig. 3. 
The variable-frequency oscillator consists of a standard multi-vibra- 
tor which has a normal frequency of approximately 60 cycles. The 
output 60 cycles from the variable-frequency oscillator drives the 50- 
watt power amplifier, which in turn powers the playback capstan 
motor. This amplifier uses two Type 807 tubes in the output stage. 

Assume for the moment that a tape has been threaded into the 
playback machine. This tape has previously been recorded by the 
equipment described in Fig. 2 at the same time that a picture film was 
made. When the playback equipment is started, the second syn- 
chronous motor will begin to operate and turn the potentiometer in 
such a direction as to increase the frequency of the variable-frequency 
oscillator. The output frequency of the 10-watt, 60-cycle amplifier 
will within approximately 0.1 sec be greater than 60 cycles, resulting 
in a rotational speed difference between the two motors which will 
very quickly turn the stator of motor number two in the opposite 
direction from that in which it first started to move, and correct the 
frequency of the variable-frequency oscillator so that the output 
derived from the tape exactly matches the power line frequency. 
When the frequency of the power feeding both of the two motors is 
identical, there is no resultant rotation of the stator of motor number 
two. This is the static condition which exists during normal play- 
back. Assume that the frequency derived from the tape was 0.1% 
low. Then the synchronous motor number one would operate at a 
slower speed than synchronous motor number two, resulting in a slow 
change in the position of the stator of number two, which in turn, 
results in a change in the variable-frequency oscillator which will 
increase its frequency and hence correct the tape speed. 

If it is necessary to cue the sound with the picture, for example dur- 
ing playback of a television show, it is possible to start the equipment 
a short time before or after the camera is started and have the sound 
track very closely match the picture. If, however, after starting, the 


picture does not synchronize with the sound, that can be very easily 
corrected by the manual speed control associated with the variable- 
frequency oscillator. If the operator finds that a correction is neces- 
sary, he will make it by turning the manual speed control until the 
meter in the lower left corner indicates zero. At this time he throws 
the automatic manual switch to "manual," and increases or decreases 
the speed of the tape to synchronize it with the picture. When the 
picture has been thus synchronized, he readjusts the meter to zero, 
throws the automatic manual switch to the "automatic" position, and 
allows the automatic correction to carry on from there. 

This equipment is designed specifically for use with any Ampex re- 
cording equipment and permits its use without requiring any modifi- 
cation of the standard recording equipment. The savings realized 
through the use of K~ m - magnetic tape for sound track recording far 
outweigh the cost of this additional control equipment, and through 
the use of magnetic tape the motion picture industry can realize even 
higher quality sound recording than it has in the past. 

Electrical and Radiation 
Characteristics of Flashlamps 



SUMMARY: Measurements of flashtube current and potential have 
been obtained using a radar synchroscope, and from these the power and 
energy supplied to the discharge for a wide range of operating conditions. 
Simultaneous observations of the time variation of the radiation in three 
spectral regions were recorded using multiplier phototubes. A lag of sev- 
eral microseconds in peak radiation behind peak power input is observed, 
the lag increasing with wavelength of the intense continuum produced by 
these discharges in rare gases. In addition to this change in quality of the 
radiation with time during the flash period, an increase in radiation efficiency 
with energy input occurs, the rate of increase being the highest for short 

THE MOST INTENSE light source commonly available is the brilliant 
flash produced by the discharge of a condenser through a gas at 
reduced pressures. The spectrum of the radiation emitted is a con- 
tinuum upon which a few emission lines are superimposed . In appear- 
ance the light is an intense white and produces an effective duplication 
of daylight illumination for photographic purposes. During the 
recent war, high current flashtubes were used extensively in reconnais- 
sance photography. More recently rapid advances have been made in 
the application of gaseous discharge flashlamps in high-speed photog- 
raphy. 1 They are commonly employed for stroboscopic work and 
have recently been used in airport runway marker systems. 

Intensities as high as 10 6 candles/sq cm have been obtained in single 
flashes in lamps where the average power input is 10 megawatts during 
the period of the flash. Light output efficiencies of the order of 50 
lumens per watt have been measured in single-flash, high-current 
discharge tubes. 2 

The present report is concerned with electrical and radiation char- 
acteristics of flash discharges in quartz tubes filled with rare gases 
having pressures in the neighborhood of 100 mm Hg. Current, 
potential, and power input to the discharge, and light output as 
measured by means of phototube multipliers, were recorded on an 
oscilloscope screen. A triggering circuit has been designed to produce 
synchronous current pulses with a repetition error less than 0.1/*sec 
(microsecond). Repetitive flashing at rates of from 1 to 60 flashes/ 
sec were used in these experiments. 

PRESENTED: April 26, 1950, at the SMPTE Convention in Chicago. 





Radiation efficiencies are found to increase rapidly with energy 
input in the visible and ultraviolet regions. In the present work 
separate radiation-time curves for the ultraviolet, visible and near 
infrared regions were obtained. In all cases the radiation reaches a 
peak several microseconds after the peak input power, the maximum 
emission occurring at progressively later times the longer the wave- 


Synchronous Pulse Generator 

The equipment used for synchronous operation of flashtubes is shown 
in the block diagram, Fig. 1. A 10-in. disc of Dow metal is driven 








3600 RPN 







4- 4- 4- 

Fig. 1. Block diagram of apparatus. 

at 3600 rpm by a Bodine Electric Co. J^o-hp synchronous motor. 
A 0.2-mm radial slit on the periphery of the disc passes light from the 
linear filament of a GE Mazda, 75-v, 4-amp movie exciter lamp to 
an RCA 1P22 photomultiplier tube to provide sharply defined current 
pulses. These pulses are amplified and used to trigger the oscilloscope 
sweep and also to initiate the discharge in the flashtube. 

A sealer circuit provides pulsing rates lower than the 60 pulses/sec 
set up hi the multiplier. It consists of six binary stages so arranged 
that any desired number may be inserted in the circuit to reduce the 
pulsing rate by a factor 2 n , where n is the number of stages used. 
Thus, in addition to the 60 pulses/sec initial rate six other rates of 
30, 15, 7.5, 3.75, 1.88 and 0.94 pulses/sec are available. In this 


manner a wide variety of flash energies may be employed, permitting 
the average power input to the flashtube to be kept low enough to pre- 
vent overheating or excessive sputtering of electrode materials. 

The flashtubes had plane parallel electrodes 17 cm apart, separated 
by a quartz envelope 16 cm long with an internal diameter of 4 mm. 
Most of the data here reported were obtained with tubes filled with 
neon or argon at a pressure of 75 mm Hg. The electrodes were con- 
nected permanently to a condenser charged to a potential less than 
the breakdown voltage (~ 4000 v) of the gases used. The discharge 
is initiated by the application of a very rapidly changing potential to 
an external " trigger" electrode located near the center of the quartz 
envelope. The action of the rapidly changing field is to produce suf- 
ficient ionization of the gas for a discharge to occur, and the condenser 
potential decreases in a few microseconds from an initial value of 
1500-3000 to a few hundred volts. The dielectric strength of the 
un-ionized gas is restored in a few tenths of a millisecond after the 
initiation of the high-current arc. A power supply is used which is 
adequate to charge the condenser to the original potential between 
recurring discharges. 

Control and Measuring Circuits 

Figure 2 gives a detailed diagram of the pulse network, the current 
shunt and the potential divider circuit arrangement, and the multi- 
plier phototube connections to the synchroscope. 

To provide high-voltage pulses for initiating the discharge, the 
amplified photo-current pulses from the sealer circuit are transformed 
to very sharp positive pulses by an RCA 884 thyratron. These pulses 
are then impressed upon the grid of a Western Electric 5D21 hard tube 
pulse tetrode, normally biased to cut off, reducing the plate imped- 
ance from a high to a very low value. The trigger electrode is con- 
nected directly to the plate of the 5D21 tube and through a one-meg- 
ohm resistor to a variable high-voltage supply. When a positive 
pulse reaches the grid the tube conducts and causes a sharp reduction 
in the potential of the external electrode from several thousand volts 
to a very low value. Observations using the fast sweep of the syn- 
chroscope indicated that the change occurs in less than 0.1 jusec, 
or at a rate greater than 10 11 v/sec. It was possible in this manner, 
using 6000 to 8000 v on the trigger electrode, to pulse successfully 
all the tubes used in this work in a perfectly consistent manner at any 
arbitrary rate for which overheating of the electrodes did not occur, 
and for observation periods of an hour or more. 

All measurements were made with a Navy radar synchroscope 






Fig. 2. Detailed 










circuit diagram. 




CL96 mfd 

2630 Vtoifs 

Fig. 3. Synchroscope 
traces of flash current, po- 
tential and photo-current. 

I 3 




Fig. 4. Variation of tube resistance with flash energy. 


Model TS-28/UPN having sweep ranges of 1-2, 10, 25 and 60 /*sec/ 
in., and provided with time markers. The sweep was triggered by 
means of a positive or negative pulse from the pulse generator circuit. 
The low-impedance input circuits were properly matched with well 
shielded coaxial cables. Care was taken to use identical cables in 
order to give correct phase relations between current, potential and 
light pulses. Direct connections to the deflection plates were made at 
a terminal strip on the back of the instrument. 

The experimental data were obtained from photographs of the syn- 
chroscope screen showing the synchronized recurrent traces of poten- 
tial, flash current and photo-currents due to the emitted radiation. 
Successive exposures on the same negative were made of all pertinent 
traces for a given set of conditions in the flash-tube circuit. These 
photographed traces of the current, potential and radiation pulses are 
thus recorded on the same negative in the correct relative phase rela- 

Measurement of tube potentials was carried out by means of a com- 
pensated and shielded high-resistance voltage divider. Space does 
not permit a detailed description of this unit and of the care taken in 
determining the correct method of its use in the flash circuit. 3 Check 
measurements showed that tube potentials read from the synchro- 
scope photographs were in error by not more than =*=5% for values 
above a few hundred volts. 

Current pulses were obtained by the use of a specially constructed 
bifilar shunt element having a resistance of .089 ohm. Such elements 
are commonly employed in measuring heavy lightning surge currents. 4 
Peak pulse current values determined from oscilloscope traces were 
checked against magnetic link measurements. The greatest differ- 
ence at high values of current, where errors due to self-induction in the 
shunt are largest, was about 8%. The average deviation in mean 
current values as determined by comparing the charges delivered by 
the condenser with the charges obtained by graphical integration of the 
synchroscope current trace amounted to d =3%. 

For radiation measurements in the ultraviolet region an RCA 1P28 
multiplier phototube was used in conjunction with a Corning 9863 
filter to give an over-all response extending from 2400 A (Angstrom 
units) to 4200 A with a peak at 3350 A. For the broad "visible" 
region an RCA 931-A multiplier was used without filter. A narrow 
visible band was obtained by using the 1P28 multiplier with a Wratten 
K-3 (No. 9) filter which limited the response to the region between 
4600 A and 7000 A. A six-stage Farnsworth multiplier with Cs-Ag-0 
cathode was used in conjunction with a Wratten A (No. 25) filter to 

292 OLSEN AND HUXFORD September 

provide a response ranging from 6000 A to 12,000 A with a peak at 
8500 A in the near-infrared region. 

Since peak light intensities ranging up to 10,000,000 lumens are 
encountered at the highest flash energies, considerable attenuation was 
required to limit the operation to the region of linear response of the 
multipliers. A fixed amount of attenuation was provided by placing 
a piece of exposed photographic film over the opening in the multiplier 
housing. For controlling the photomultiplier currents during a series 
of measurements ranging from low to high peak light intensities a 
neutral Eastman Kodak filter having calibrated sectors was employed. 
In the present study light intensities are expressed in arbitrary units 
in each of the three spectral regions. 


Figure 3 shows typical photographs of current, potential and photo- 
current traces for flash discharges in neon and argon. Exposure times 
of from 40 to 60 sec were required, so that at the repetition rate of 
0.94/sec, each trace represents a large number of recurring discharges. 
The sharpness of these composite traces indicates the very precise 
manner in which the discharge repeats itself. 

In these pictures the photo-current trace was obtained with the 
931-A multiplier phototube and denotes radiation emitted in the 
broad visible range. Peak light intensities lag behind peak currents 
by about 5 jusec at low potentials. This lag decreases as the input 
energy per flash is increased. The main objective of the present in- 
vestigation was to study in detail the changes in intensity and quality 
of the radiation with energy input and with time during the flash 

Flashtube Resistance 

Following the method of Murphy and Edgerton, 5 a quantity is used 
which we have called the "tube resistance." It is defined by the 

Rt = V m /I mj (1) 

where I m is the value of maximum flash current, and V m is the poten- 
tial difference between the electrodes at the time of peak current. 

The variation of R t with energy input per flash in neon and argon 
flashtubes of identical size and gas pressure is shown in Fig. 4. 

It is found that the flash current decays in an exponential manner 
according to the equation 




i = /, 

where the values of X are given, within a mean error of 

20%, by the 


X = 1/RtC . 

This shows that the rate of discharge of the capacitance, C, is very 
closely that to be expected in an RC circuit in which R is equal to the 
"tube resistance" defined in Eq. (1). 

Energy Supplied to the Discharge 

In order to determine radiation efficiencies, the fraction of the 
capacitor energy actually consumed in the discharge must be known. 
The power delivered to the tube as a function of time is calculated 



10 15 20 


Fig. 5. Computed power-time curve. 

from the product of simultaneous values of current and potential. 
An example of a power curve obtained in this way from replotted 
synchroscope traces is shown in Fig. 5. Graphical integration of the 
power curve yields, for each flashing condition, the energy delivered 
per flash to the discharge. 

Energy calculations were carried out for one neon tube and one 
argon tube; three capacitors were used with potentials ranging from 
1000 to 3000 v. The results appear in Fig. 6, where energy in watt 
seconds per flash is plotted against peak current values, both scales 
being logarithmic. Within the limits of experimental error, energy 
per flash is proportional to peak current, for a constant value of 
capacitance and variable voltage. 




The measured energy consumed by the discharge differs from the 
energy of the charged capacitor by a fraction of one per cent at low 
potentials, and increases up to 15% at high potentials. The discrep- 
ancy will be much greater in circuits where lead wires of minimum 
length are not employed. In the present experiments the losses in 
leads and condensers correspond to those in a resistance of the order of 
0.1 ohm in addition to the .089-ohm resistance of the current shunt. 



2 4 6 10 


4 6 10 


Fig. 6. Peak current vs. measured energy per flash. 

Radiation Intensity vs. Input Energy 

Measurements of the intensity of emitted radiation averaged over 
the entire flash period, determined either from single flashes or by 
using repetitive flashing, have shown that the following relation holds 
approximately : 

= cW n 


Here $ is the average (integrated) intensity, c is a constant for a given 
tube, W is the energy of the charged condenser, and n ^ 1, its value 
depending upon the filling gas and spectral quality of the radiation 
reaching the phototube. 

Examples of such results are shown in Fig. 7 for GE FT-14 xenon- 
filled flashtubes. Curves (a), (b) and (c) were obtained in this labora- 
tory at low pulse rates (~ 10 pulses/sec), using widely different values 
of capacitance. Readings were made using a 931-A phototube multi- 
plier, and light intensity and flash energy are expressed in arbitrary 
units. Curve (d) is taken from results published by Edgerton, and the 




Fig. 7. Radiation in- 
tensity vs. capacitor en- 
ergy for GE xenon flash 

10 s 


- - 


-t * 


> fl y 


/ ' 

/^ "K- 




>l f 







































5 2pf. 











./ // 









[a) 1000 
[b) 1000 

to 2 
to 2 









= .54 


c) 1000 

to 2 




2] ^^ 









d) FT-14 







L c 












le ii 


Vatt-Sec for 





IO/ 10* 



Fig. 8. Radiation 
intensity vs. peak cur- 
rent and flash energy 
for argon and neon. 

2 46 l 


2 46 10 watt-sec. 


296 OLSEN AND HUXFORD September 

integrated radiation is given in lumen-seconds as a function of the con- 
denser energy per flash in watt-seconds. The value of n in all of 
these plots is about 1.5. 

If an ultraviolet filter is used, light in the near ultraviolet region can 
be recorded. The FT-14 Tubes have a pyrex envelope and protecting 
outer chimney of pyrex, so that short-wave ultraviolet light is not 
transmitted. For this radiation the plot of Fig. 7 (e) was obtained, 
the slope of which yields the value n = 1.8. A large number of meas- 
urements on similar xenon-filled tubes of pyrex yielded a mean value 
ofn = 1.77. 

Additional measurements using a Farnsworth multiplier and infra- 
red filter showed that n = 1.0 =*= 0.1 for. xenon flashtubes. Hence, in 
the near infrared spectral region the efficiency is constant, the light 
intensity increasing linearly with flash energy. 

The results of measurements of the integrated radiation emitted 
by argon- and neon-filled quartz tubes made in this laboratory, using 
a 931-A multiplier with no filter, are shown in Fig. 8. The mean 
value of n for argon is 2.15; for neon, 2.3. When measurements 
were carried out hi the three spectral regions with the photomultiplier 
tubes and filters, described under Apparatus above, the results shown 
in Table I were obtained for radiation integrated over the entire flash 

TABLE I. Value of n in the equation S = cW n 

Mean Values of n 


Spectral Range 




2400- 4200 A 




Visible. . . . 

. . .4600- 7000 A 


6000-12,000 4. 

Phase Lag of Light With Respect to Flash Current 

Ultraviolet radiation peak intensities and to a lesser extent peak 
visible light intensities in both neon and argon were found to increase 
more rapidly with flash energy than do the mean values of the inte- 
grated light intensities. In addition, these peak values occur earlier 
in the flash period the shorter the mean wavelength of the spectral 
region. This is indicated very clearly in the synchroscope traces of 
Fig. 9 for photo-currents in the three multipliers for identical flashing 
conditions. In these plots photo-currents representing light inten- 
sities in the three regions of the spectrum have been plotted so that 
peak values are nearly the same. Light intensities are comparable 




Fig. 9. Photo-current traces 
in three spectral regions. 


? e 

Eu. o 






C-I.S 6|il 


V.* ISOOv. 

CM. >6pf. 

4 8 12 16 20 24 28 32 

only for any one trace; actually the intensities are highest in the ultra- 
violet, nearly as high in the visible and much lower than either of 
these in the infrared region. 


There are two aspects of the results obtained in this work which are 
of importance in the use of flashtubes for photographic and illumina- 
tion purposes, and which require consideration in a detailed analysis 
of the discharge process. The first is the rapid increase in radiation 
efficiency with energy input, the increase being greatest for short 
wavelengths. A study of flash discharges in capillary tubes reported 
by Hahn and Finkelnburg 7 showed that the intensity of the continuum 
at all wavelengths increased as the square of the current density for 
densities greater than about 70,000 amp/sq cm. These authors 
believe that this continuum is largely due to the retardation of elec- 
trons in the fields of the ions in the discharge plasma (Bremscon- 

Observations in this laboratory of the nature of line spectra emitted 
by flash discharges show that, in the decaying portion of the radiation 
pulse, recombination of electrons and ions is occurring at a rapid rate. 
This process, involving "free-bound" transitions between electrons 
and ions, must also contribute to the observed continuum. The fact 
that radiation intensities vary not simply as 7 max 2 (where 7 max is the 
peak current), but as / max n > where n varies from 1.2 to nearly 3.0 
depending on wavelength, indicates that probably the exciting elec- 


irons undergo considerable change in velocity distribution during the 
flash period thereby modifying the simple quadratic relationship 
predicted for Bremsstrahlung and recombination continuua. 

The second aspect to be noted is the fact that peak radiation occurs 
at different times depending on the spectral region observed. This 
phenomenon suggests the conception that the densely ionized plasma 
radiates as a "gray body" and exhibits a spectral maximum which is 
temperature dependent. Early in the discharge cycle the radiation 
peak is in the ultraviolet region indicating a high electron tempera- 
ture. With cessation of current flow the plasma cools rapidly due to 
radiation, conduction and convection, and due also to the rapid 
expansion of the hot gases in the discharge column. As the mean 
electron temperature falls, the radiation peak shifts to longer wave- 
lengths much as in the case of incandescent solids. Observations of 
electron excitation temperatures, both as a function of time during the 
flash and as a function of the energy supplied to the discharge, are 
being carried out in this laboratory in an attempt to correlate these 
two aspects of the condensed flash discharge. 


1. "High-speed photography," Jour. SMPE, vol. 52, pp. 5-116 of Part II, Mar. 


2. J. H. Aldington, "The high intensity flash discharge tube," Endeavour, vol. 7, 

p. 21, Jan. 1948. 

3. A description of a similar device has recently been published: H. J. White, 

"Transient response of high voltage resistance dividers," Rev. Sci. Instr., 
vol. 20, pp. 837-839, Nov. 1949. 

4. John H. Park, "Shunts and inductors for surge-current measurements," J. 

Res. Nat. Bur. Stand., vol. 39, pp. 191-272, 1947. 

5. P. M. Murphy and H. E. Edgerton, "Electrical characteristics of stroboscopic 

flash lamp," /. App. Phys., vol. 12, pp. 845-855, Dec. 1941. 

6. H. E. Edgerton, "Photographic use of electrical discharge flash tubes," /. Opt. 

Soc. Amer., vol. 36, pp. 390-399, July 1946. 

7. O. T. Hahn and W. Finkelnburg, "Continuous Bremsstrahlung and recombi- 

nation radiation of electrons in the gas discharge plasma," Zeits.f. Phys., vol. 
122, pp. 37^8, 1944. 

The Cine Flash 

A New Lighting Equipment for High-Speed 
Cinephotography and Studio Effects 





SUMMARY: A new form of portable lighting equipment is described which 
has been designed especially to meet the needs of the high-speed cinephotog- 
rapher who is always faced with the difficulty of obtaining sufficient light. 
Two compact-source mercury cadmium lamps are operated in series at their 
normal wattages of 1 kw, and are then flashed at 3, 5 or 10 kw for 5, 2 or 1 
sec. The equipment consists of a control unit and two lightweight lamp- 

The light output is sufficient for color photography at speeds up to 3000 
frames/sec, or for black-and-white photography with small lens apertures to 
give considerable depth of focus. The flash may be triggered from a micro- 
switch or from a camera switch. The steady light output from the lamps is 
sufficient to arrange and focus the subject. 


FOR THE SCIENTIFIC PHOTOGRAPHER the high-speed cinecamera is 
a valuable and potent tool enabling him to study extremely rapid 
motion with ease and certainty. 1 Projection of the cinefilm at a low 
speed enables the apparent movement to be slowed down so that it 
can be followed by normal vision. Using this technique, or that of 
projection frame by frame, a detailed analysis of the motion can be 
made and data such as the velocity, acceleration and position of the 
object under observation at various instants may be obtained. To 
the designer of machinery in particular, the technique of high-speed 
cinephotography has proved invaluable; without it many of his more 
difficult problems would still remain unsolved. Also, in the science 
of ballistics many advances may be attributed directly to the use of 
the high-speed cinecamera. 

The chief applications of high-speed cinephotography lie in the 
fields of science and industry. 2 Modern high-speed cameras used in 
the majority of these applications generally operate at speeds up to 
3000 frames/sec. At such speeds the exposure time is extremely 
short: only Ksooo sec at 3000 frames/sec. Such a short exposure 
PRESENTED: October 13, 1949, at the SMPE Convention in Hollywood. 


300 BOURNE AND BEESON September 

time necessitates an extremely high illumination; for example, even 
with a very rapid film such as Kodak Super XX, an illumination of 
some 10,000 ft-c (foot-candles) will be required to expose an average 
subject with an aperture of //2.8 at 1500 frames/sec. Owing to the 
slow emulsion of color film, it is even more difficult to provide suf- 
ficient illumination for high-speed color cinephotography, and the 
photographer is continually faced with the problem of obtaining 
enough light to allow a small enough lens aperture to give adequate 
depth of focus. Fortunately in many applications of high-speed 
cinephotography the subject is small in size and an illuminated 
area of only 6 to 12 in. in diameter is often sufficient. It is fortunate 
too that at these high speeds long periods of exposure are rarely neces- 
sary and useful total exposure times generally lie between 1 and 5 
sec. It thus appears that a lamp providing up to 100,000 ft-c over 
an area of not more than 1 ft in diameter, for a duration up to 5 sec, 
would satisfy many of the needs of the high-speed cinephotographer. 
Because many of the applications are in factories, laboratories and 
industrial organizations, light weight, robustness and portability 
are other essential requirements for the equipment. 

The difficulty of obtaining adequate illumination was stressed in 
this Society's Symposium 3 which included a number of valuable papers 
dealing with various aspects of the subject. In one paper dealing with 
lamps for high-speed photography, the requirements of the ideal light 
source were outlined and the paper then described the various methods 
which are being used for providing illumination. Each of these 
methods has certain limitations. 

In the past, photographers have generally used standard film- 
studio incandescent spotlights for lighting the subject. For ex- 
ample, one M-R 414 Fresnel-lens spotlight with a 5-kw incandescent 
lamp produces approximately 10,000 ft-c over a 12-in. diameter spot 
at a distance of 5 ft. It is however, difficult to group enough spot- 
lights closely together to obtain sufficient illumination; again, 
heating of the subject is extremely severe so that special means of 
cooling are often necessary. Another interesting but expensive 
method of lighting used in the United States 4 is to produce a short 
continuous flash by successive firing of a number of aluminum foil 
flashbulbs mounted on a rotating disc and passing in turn in front of a 
mirror. Highly loaded, short-life filament lamps intended only for 
intermittent burning have also been employed successfully. 5 The 
electronic flashtube is another light source which provides an ex- 
tremely high light intensity but the duration of the flash is only a few 
microseconds so that, while it is eminently satisfactory for taking 




Fig. 1. 1-10-Kw pulse-type compact-source lamp. 

single high-speed still photographs, it is not suitable for cinephotog- 
raphy. 6 

None of the above methods meets all the requirements of the high- 
speed cinephotographer. The development of the discharge lamp 
has made possible a new means of obtaining a sufficiently high light 
output by flashing the lamp at a high overload, and special equip- 
ment has been designed to utilize this principle. 

Certain technical problems necessitate ultra-high-speed photog- 
raphy at speeds far higher than 3000 frames/sec. Special cameras 





PO OJ -k 
D 3 w 


p / 













p i 







F l>i 



. 1 















* IS 








4000 5000 

Fig. 2. Spectral distribution of mercury cadmium lamp. 


are required, together with still higher lighting intensities. In this 
field, too, the new lighting unit described in this paper should be more 
effective than other sources hitherto available. 


The compact-source mercury vapor lamp which is available in 
sizes from 250 to 10,000 w in England is now also becoming known in 
the United States. 7 A typical lamp, which is shown in Fig. 1, con- 
sists of a spherical transparent quartz bulb containing mercury and a 
rare gas filling. Two tungsten electrodes between which an arc 
operates are sealed into the bulb and the current is led in by molyb- 
denum foil seals diametrically opposite to one another. Brass caps 
are fitted at tne end of the seals. Connections are made through a 
flexible lead to the anode and through the special base to the cathode 
of the lamp. The light source itself is small in size and has a bright- 
ness of approximately 35,000 candles/sq cm with a luminous efficiency 
of 45 to 50 lumens per watt. At the normal rated power the life of 
the lamp averages 500 hr. An important feature of the compact- 
source lamp is that it may be operated for short periods at ratings 
greatly in excess of its normal power, and under these pulsed condi- 
tions it produces a correspondingly high light output. 8 

The radiation from a compact-source mercury vapor lamp is dis- 


continuous and consists mainly of yellow, green and blue lines super- 
imposed on a relatively weak continuous spectrum. Owing partic- 
ularly to the deficiency in red content this light source produces a 
distorted rendering of colors but a considerable improvement in the 
color rendering can be obtained by the addition of cadmium in the 
discharge. 9 The discharge in cadmium vapor produces a powerful 
red line in the spectrum, while the gaps in the blue-green region of the 
mercury spectrum are filled by additional lines, so that the color of 
the radiation is better balanced. The color rendering also improves 
with increased current density in the arc so that the higher the 
wattage of the lamp the better is the color rendering. This improve- 
ment is due partly to broadening of the lines and partly to an increase 
in the amount of continuum at the higher current density. Spectral 
distribution diagrams of the mercury cadmium lamp operating at 1 
kw and at 10 kw are shown in Fig. 2. These diagrams show that 
gaps still remain in the spectrum, but in spite of these gaps practi- 
cal tests have shown that mercury cadmium compact-source lamps 
operating at powers above 2.5 kw give a satisfactory rendering of 
color with Kodachrome Daylight and similar emulsions. 


A portable equipment using the principle of flashing a compact 
source lamp has been designed specifically to meet the illumination 
requirements of the high-speed cinephotographer. One equipment, 
known as the M-R 356 Cine Flash, illustrated in Fig. 3, consists of 
two lamp heads mounted on stands, and a control unit containing 
the ballast resistance and other components for operating the lamps. 

The lamphouse, an interior view of which is shown in Fig. 4, is of a 
light sheet-metal construction. It contains a compact-source lamp 
mounted along the axis of a paraboloidal mirror and is fitted with a 
front diffusing glass. This optical arrangement gives a high light 
collection efficiency combined with uniform distribution. The 
mirror is made of metal so that it is robust and not liable to damage 
during transport of the equipment. By releasing the locking screws 
at the rear of the lamphouse, the lamp may be removed through the 
front of the housing for transport. The high-frequency choke re- 
quired for the impulse striking circuit is mounted behind the mirror 
in the lamphouse. 

The lamp is mounted in a prefocused holder and the focal position 
is set normally to produce a spot 10 in. in diameter at a distance of 4 ft 
from the unit with an illumination constant to within =^15% over 
its area. A fixed rather than variable focus ensures that the light 
output from the unit is constant so that the photographic exposure 




Fig. 3. Cine Flash Equipment. 

can be repeated accurately by setting the lamp at predetermined 
measured distances from the subject. In cases where it is necessary 
to cover a wider area with the lamps, the spread can be increased 
approximately threefold by using a front glass giving greater diffusion, 
but the light intensity is reduced to approximately 34 o in this case. 
A polar distribution curve of the equipment operating at 1 kw with 
the standard diffusing glass is shown in Fig. 5. 

The light source is a color-modified mercury cadmium compact- 
source lamp normally rated at 1000 w but it differs from the standard 
type in that it has been designed specifically for this flashing service. 
The bulb is similar in design and dimensions to a standard lamp but 
special massive electrodes visible in Fig. 1 are provided to withstand 




Fig. 4. Lamphouse with front glass removed. 

the heavy overload conditions of flashing without fusing and con- 
sequent blackening of the bulb. A special construction giving good 
heat conduction is used to prevent melting of the electrode tips during 
the pulse. The very high current density in the flash condition en- 
sures that the color of the radiation is good. As repeated flashing 
causes a further slight increase in the red content due to additional 
evaporation of cadmium at the higher bulb temperature, the color of 
the radiation does not stabilize completely until the lamp has been 
flashed several times in succession. Before taking a color photo- 
graph with this equipment, to ensure the best color rendering, it is 
therefore advisable to stabilize the color of the radiation by flashing 
the lamp several times. Alternatively the bulb may be preheated 
by a continuous moderate overload of some 50% for 30 sec. The 
overload is applied by depressing a push button; when this button 




Y* 2000 





S 1000 




























s s 



3 10 5 5 10 IJ 

Fig. 5. M-R 356 Cine 
Flash polar distribution; 
lamp operating at 1 kw, 
measured at 10 ft with 
normal diffuser, 18,430 lu- 
mens from each lamp. 

is pressed a red warning lamp lights up to indicate that the lamp is 
being overloaded. 

The light output from a discharge lamp follows changes in the 
current through it almost instantaneously. For example when a 
compact-source lamp operates on alternating current at 50 cycles/sec, 
the light falls to 4% of the maximum at the end of each cycle. In 
order to prevent cyclic changes in exposure of the film it is therefore 
necessary to operate the lamp on a smoothed d-c supply and the 
ripple voltage should preferably not exceed 10% of the supply voltage. 

The control unit is designed normally for operation from a 200- 
to 250-v d-c supply. When no d-c supply mains are available, the 
equipment must be operated either from a three-phase rectifier unit 
or from a mobile d-c generator. 


The first type of equipment to be built consisted of a control unit 
with one lamp head. This equipment was demonstrated at the 
Royal Photographic Society on January 13, 1949, and subsequently 
at the British Kinematograph Society. Generally, however, two 
light sources are necessary for photography to obtain the necessary 
light distribution and modeling. Later equipment was therefore 
redesigned to operate two lamps in series with a single control unit 
thus enabling the light output to be doubled for the same current 
from the supply. 

A mains voltage selector link on a panel at one end of the control 
unit is provided for setting the equipment for the particular supply 
voltage which can be read on the panel voltmeter. This ensures that 


the lamps will always operate at their correct wattage. The lamps 
are started by a high-voltage impulse circuit contained in the control 
unit, the impulse being applied to the lamp through an insulated high- 
tension cable. The circuit produces a steep-fronted impulse of 
approximately 15 kv and will reignite the discharge even if the lamp 
has not cooled down completely so that inconveniently long delays 
usually associated with high-pressure mercury vapor lamps are re- 
duced. The arc is ignited by opening and closing a spring-loaded 
striking switch on the control unit. This switch discharges a con- 
denser through the primary of the pulse transformer and produces 
a high-voltage high-frequency impulse across the secondary winding. 
A high-frequency choke in series with one of the lamps and mounted in 
the lamp housing prevents the impulse from being short circuited 
by the low-impedance path through the supply mains. 

Immediately the lamps strike, the lamp voltage indicated on the 
voltmeter falls to approximately 20 v and as they warm up the lamp 
voltage rises. When the total lamp voltage reaches approximately 
100 v the starting resistance which limits the starting current should 
be short circuited by throwing the starting switch over to the "Run" 
position. After this the lamps will rapidly reach their final operating 
condition. Interlocking of the starting and striking switches pre- 
vents accidental striking of the lamp if the starting switch is in the 
"Run" instead of the "Start" position. The run-up process takes 
approximately 10 min. 

The lamps are operated normally at 1000 w with a series ballast 
resistance. To flash the lamps a section of the resistance is short 
circuited by a contactor, the duration of the flash being predeter- 
mined by a resistance-condenser timing circuit. Selector switches 
controlling the power and duration of the flash are interlocked to 
give flashes of 3 kw for 5 sec, 5 kw for 2 sec or 10 kw for 1 sec, as 
required. The duration of the flash and the power in it are limited 
by temperature considerations; at 5 and 10 kw, melting of the elec- 
trodes is the limiting feature while at 3 kw the limit is set by heating 
of the quartz bulb. After the lamps have been flashed, an interval 
of at least 30 sec must elapse before they are flashed again, in order 
to prevent damage due to overheating. This interval is provided 
automatically by a timing circuit which must be reset manually by a 
push button before the lamp can be flashed a second time. This 
circuit cannot be reset until the necessary 30-sec interval has elapsed. 

The desirability of preheating the lamps before taking a color 
photograph has already been mentioned. This is done by pressing 
the "Preheat" button in the control unit. Part of the ballast resist- 

308 BOURNE AND BEESON September 

TABLE I. Characteristics of 1-10-kw Pulse Compact-Source Lamp 

Over-all length 245 mm 

Bulb diameter 45 mm 

Arc length 6 mm 

Lamp wattage 1, 3, 5, 10 kw 

Lamp current, approximate 15, 40, 70, 125 amp 

Lamp voltage, approximate 70 v 

TABLE II. Illumination and Spot Size Given by Cine Flash 



Glass Giving 

Wider Divergence 

Ilium, at 

Dia. of spot 

Ilium, at 

Dia. of spot 


center of spot, 

to 70% of 

center of 

to 70% of 



max., in. 

spot, ft-c 

max., in. 














































The above figures show the illumination and area covered by each lamp at 
various distances at 10 kw. The illumination at other wattages may be taken as 
proportional to the wattage. 

ance is thereby short circuited and the required overload is applied 
to the lamps for 30 sec during which time the bulbs reach the correct 
temperature and the color stabilizes. The lamps should then be 
flashed within the next 30 sec ; if they are not flashed within that time 
the preheating procedure should be repeated to offset the cooling 
which has occurred during the waiting period. Preheating is not 
necessary for black-and-white photography. 

The flashing circuit can be operated either by a push button on the 
control unit or from an external switch connected by flexible leads 
to a socket in parallel with this push button. Momentary closure by 
a microswitch will operate the equipment and the flash can be initi- 
ated by a normal built-in camera switch such as that used on the 
Eastman high-speed camera. The time required to initiate the. flash 
is limited chiefly by the speed of closing of the contactor; it is only a 
few milliseconds. Two sockets in parallel on the control panel enable 
several units to be flashed synchronously from the same switch if 

In normal operation at 1000 w, the lamp current is 15 amp. When 




TABLE HI. Light Output From Cine Flash Unit 

Light intensity 
at center of 

12-in. circle at 


Supply Current 

4 ft from unit, 


230-v, d-c, 





amp per ph. 

Cine Flash 

At 1-kw rating 





At 3-kw 





At 5-kw 





At 10-kw 





M-R 414, 5-kw 

Incan. Studio 





The figures above give the light output from each lamp with the normal diffusing 
The values can be doubled if the spots from the two lamps are arranged to 
overlap one another. 

the lamp is flashed at 10, 5 and 3 kw, the respective values of the 
current are 125, 70 and 40 amp, with corresponding durations of 1, 
2 and 5 sec. Although the peak operating current of the equipment 
is very high, the duration of the surge is quite short. Even so, there 
may sometimes be a difficulty in providing these high peak currents 
in locations where the power supply is limited and it is therefore neces- 
sary before using the equipment to check that the supply mains are 
fused adequately. 

Owing to the high light intensity given by this unit, exposure is 
best judged by making practical tests with the lamp at various dis- 
tances from the subject. Once the correct exposure for a certain 
distance has been found there will be no difficulty in repeating the 
results. With the normal setting of the lamp the light intensity at 
the center of a 10-in. spot at 4 ft from the unit is approximately 15,000 
ft-c at 1 kw, 45,000 ft-c at 3 kw, 75,000 ft-c at 5 kw and 150,000 ft-c 
at 10 kw. The exposure may also be measured with a meter with the 
lamp operating steadily at 1 kw and then decreased in proportion to 
the power in the flash. 

Table I summarizes the chief optical and electrical characteristics 
of the lamp. Table II, which shows the light output and size of the 
spot at various distances from the lamphouse, will be found useful 
in estimating the exposure at other distances. At maximum power, 
the light intensity should be sufficient for taking a film at 1500 frames/ 
sec with a fast emulsion and an aperture of //ll. With Kodachrome 
film the unit should give sufficient light when flashed at 10 kw for 
photography at 3000 frames/sec at a distance of 4 ft with an aperture 




Fig. 6. Rectifier and Cine Flash Unit. 

of //2.8. Table III shows the current consumption of the Cine 
Flash Unit. 

For operation from an a-c supply, when no d-c supply is available, 
a three-phase mercury vapor rectifier has been designed. This is 
mounted in a mobile framework which is also designed to carry the 
Cine Flash Unit on top of it. The rectifier and control unit are 
illustrated in Fig. 6. A smoothing circuit is built into the rectifier 

One of the first successful high-speed films made in Kodachrome 




Fig. 7. Control box for lightning effects. 

has recently been taken by Kodak Ltd. with the M-R 356 Cine Flash 
equipment. The subject of this film is the beating of an egg and the 
fall of a lamp bulb filled with colored paints. The subject was illum- 
inated by three lamps arranged to flash in synchronism when the 
camera reached full speed. The film was made at 2500 pictures per 
second with a lens aperture of //2.7. Two lamps were used for front- 
lighting and one for back-lighting at a distance of 4 ft from the 

Equipment of this type will no doubt prove useful in many photo- 
graphic fields other than those of cinephotography. Scientific 
applications in which this equipment should find an immediate use 
include wind tunnel illumination, illumination for Schlieren equip- 
ment, projectile photography, underwater photography of projectile 
explosions or other phenomena. Another application of Cine Flash 
equipment is in lithographic printing. 

The application of this equipment in the film studio appears to be 
limited to effects lighting as the duration of the flash is far too short 
for normal cinephotography. For example, the addition of another 
small control unit shown in Fig. 7 enables the unit to be used for pro- 
ducing artificial lightning or flashing effects for film studio, television 
or for theaters. This is done by dividing up the duration of the flash 
into 16 equal intervals, each of which corresponds approximately 


to two frames of the film. The number of intervals or flashes may be 
chosen to produce the required effect by opening or closing any 
individual switches in the bank. If the equipment is then set to 
give say, a 10-kw, 1-sec flash, this flash can be broken up to simulate 
lightning, gun flashes or other effects. The residual light in the 
standard equipment is approximately Ko f the maximum which 
is too high to give a good effect on the film. To produce the best 
effect the residual light can be reduced to approximately J^o of the 
maximum by inserting an additional resistance to underrun the lamp 
considerably before the flash is produced. If no residual light what- 
ever is required, a shutter can be used on the front of the lamp which 
is opened just before flashing the lamp. This method has the ad- 
vantage that once the nature of the flash de'sired has been found by 
trial, the flash may be repeated as often as required. 

The original work on the lamp development and its application was 
carried out in the Research Laboratory, the British Thomson- 
Houston Co. Ltd., Rugby, by the authors. The practical equipment 
described in this paper has been developed and built in the Mole- 
Richardson (England) Ltd. Experimental Dept. The authors wish 
to make acknowledgments to L. J. Davies, Director of Research, 
British Thomson-Houston Co., and to Mole-Richardson (England) 
Ltd. Acknowledgments are also due to the British Thomson- 
Houston Co. for the photographs which are Figs. 1 and 2. 


1. E. D. Eyles, "Some application of high-speed photography," Phot. Jour., 

vol. 83, pp. 261-265; July 1943. 

2. E. D. Eyles, "High-speed photography and its application to industrial 

problems," Jour. Sc. Instr., vol. 18, pp. 175-184; Sept. 1941. 

3. High-Speed Photography, a symposium, Jour. SMPE, Part II, pp. 1-129; 

Mar. 1949. 

4. H. M. Lester, "Continuous flash lighting an improved high-intensity light 

source for high-speed motion picture photography," Jour. SMPE, vol. 45, 
pp. 358-369; Nov. 1945. 

5. R. E. Farnham, "Lamps for high-speed photography," High-Speed Photog- 

raphy, a symposium, Jour. SMPE, Part II, pp. 35-41; Mar. 1949. 

6. H. E. Edgerton and K. J. Germeshausen, "Stroboscopic-light high-speed 

motion pictures," Jour. SMPE, vol. 23, pp. 284-297; Nov. 1934. 

J. N. Aldington, "The electric discharge lamp," Trans. I.E.S. (London), 

Bright Light Sources, Part 2, pp. 11-39; Feb. 1946. 

7. H. K. Bourne, Discharge Lamps for Photography and Projection, Chapman 

and Hall, London, England, 1948. 

8. E. J. G. Beeson, "A high-intensity light source for high-speed kinematog- 

raphy," Phot. Jour., vol. 89B, pp. 62-67; May-June 1949. 

9. F. E. Carlson, "New developments in mercury lamps for studio lighting," 

Jour. SMPE, vol. 50, pp. 122-138; Feb. 1948/ 

A New Heavy-Duty 
Professional Theater Projector 



SUMMARY: The paper describes the new Simplex X-L 35-mm projector 
mechanism which is now in production. High lights of the improvements 
are reduction in mechanical load on the gear train, improved lubrication, 
new lens mounts, more finger room for threading, a direct viewing telescope 
focusing device and over-all design simplification. 

Driving Mechanism. Figure I is a general view of the complete 
mechanism. The main-drive gear assembly is an extremely simpli- 
fied vertical unit operated in sealed ball bearings (Fig. 2). This ball 
bearing construction, which is used throughout, together with the 
direct high-speed drive, effects a reduction in mechanical load over 
past practice of approximately 66% at start and approximately 80% 
while running. Inasmuch as excess mechanical load both at starting 
and running is the cause of the majority of projector shutdowns this 
improvement is of particular significance. 

The entire driving mechanism and the gear train are housed in an 
oil- tight enclosure and are visible at all times through a large trans- 
parent window which may be easily removed. The wide-face, heavy- 
duty type of gears are few in number and will require little, if any, 
attention. All high-speed shafts are equipped with ball bearings, 
and for added protection both upper and lower sprocket shafts are 
fitted with Oilite bearings. 

Lubrication. Figure 3 shows the Spray-O-Matic lubrication system 
used. The entire area of this sealed-drive compartment is sprayed 
continuously by a fine film of oil which reaches every drive unit with- 
out allowing a drop to leak through to the film. The oil feed unit is 
simplicity itself comprising a high-speed pump, a filter and a pipe. 
An oil gage fitted with a drainage petcock indicates the oil level. A 
change of oil is indicated approximately every eight to twelve months. 

Intermittent Movement. The intermittent movement has been re- 
designed and the flywheel has been mounted directly on the cam- 
shaft thereby eliminating the intermediate gears to give quieter 
operation and lowered maintenance costs (Fig. 2). The entire move- 
ment may be removed from the nonoperating side of the mechanism. 
In order to assure parallel assembly the cam pin is ground to its close 
PRESENTED: April 27, 1950, at the SMPTE Convention in Chicago. 





Fig. 1. Complete mechanism of the Simplex X-L 35-Mm Projector. 

tolerance after assembly to the cam. The position of the cam pin 
with relation to the cam ring is slightly adjustable, thus providing 
simplicity of assembly and replacement. 

Continuous lubrication of all parts of the movement is obtained 
through a separate pump comprising a pair of gears driven from the 
camshaft which force oil through the intermittent housing. 

Shutter. One of the most important design improvements is the 
new single-cone type of shutter assembly which is located a little 
more than 1 in. from the aperture, thereby providing an increase in 


illumination, compared with the cumbersome front and rear shutter 
assemblies (Fig. 4). The new shutter is maintained in correct timing 
by means of a sliding helical spline on the shutter shaft, there being 
no axial displacement between gears; thus gear noise is greatly 
reduced. A travel ghost adjustment knob is conveniently located 
at the top of the main frame. 

Framing. The framing device shown in Fig. 2 has convenient 
handles which protrude from the side of the case and are located so 
that the picture may be framed from either side. 

Lens Mount. Figure 1 shows the new lens mount made to hold 
accurately new-type lenses up to and including 4 in. in diameter and 
having speeds as fast as //1.6. This is of particular importance in 
theaters with long throws and in drive-in theaters when lenses of 
focal lengths greater than 5 in. are required. 

Quick, precise focusing of the lens is simplified by means of the 
unique Screenoscope device which is essentially an eight-power 
prismatic telescope mounted above the lens mount (Fig. 1). With 
the Screenoscope the projectionist may observe a highly magnified 
section of the screen and accomplish exact focus without eyestrain. 
As a matter of fact, obtaining a sharp and large focus of the tiny 
holes in the screen is easily and readily accomplished. 

Spot Sight Box. A large eye-protecting viewing glass properly 
located for easy vision so the projectionist may readily observe the 
light spot on the film aperture replaces the conventional small spot 
sight box (Fig. 1). 

Change-over. An instant-acting zipper type of change-over unit is 
part of the mechanism and is mounted above the shutter guard 
housing as shown in Fig. 1. The dowser blade is positioned between 
the arc lamp and shutter to protect the shutter blades against Avarpage 
and burning. 

Threading Compartment. The operating side of the new mechanism 
is provided with increased "finger room," thus reducing the problem 
of threading in the film and affording extremely easy operation. A 
threading lamp lights automatically when the door is opened and 
additional illumination is provided in the door itself. The readily 
removable film trap and gate are equipped with long confining 
film guides and adjustable tension shoes. Means are provided for 
easily threading in frame by the incorporation of an additional aper- 
ture in the upper section of the film trap just below the guide rollers, 
and an indicator is provided on the outward bearing arm of the 
intermittent movement to signify when the movement is in the 
locked position. The interior of the operating side of the mechanism 

Fig. 2. Main drive assembly. 


Fig. 3. Spray-O-Matic lubrication system. 

Fig. 4. Shutter assembly. 

Fig. 5. Automatic safety trip. 



is finished in white porcelain enamel and all corners are rounded to 
eliminate the possibility of dirt accumulation. While the film trap 
and gate assemblies follow closely the design of the E-7 Simplex 
mechanism assemblies, improvements have been incorporated which 
reduce the possibility of heat transference to the aperture plate. 
The adjustable gate shoe tension has been improved and a push- 
button gate closing means provided. 

Automatic Safety Trip. An improved automatic safety trip is 
provided which will drop the fire shutter should a patch part above 
the intermittent sprocket (Fig. 5) . 

24-Tooih Sprockets. An important new design feature is that both 
upper and lower sprockets have 24 teeth, 8 more than the conventional 
type, and they operate at only 240 rpm, a reduction in speed of 33J% 
over ordinary sprockets. 

Cooling. Some cooling is obtained for the aperture and film gate 
by means of air drawn through an opening behind the shutter housing, 
forced past the film trap and discharged through openings on the 
operating side of the equipment so that a constant supply of cool air 
to the film trap is available at all times when the mechanism is in 

Upper and Lower Magazines. Both magazines are considerably 
deeper than usual, to accommodate bent exchange reels. The upper 
magazine is equipped with an observation lamp and a large porthole 
so that the remaining footage may be readily observed. Also, 
a well-designed film valve is provided by means of which, through the 
addition of a large flanged roller, the film path is maintained in correct 
alignment with the upper sprocket and scratching of the picture area 
or sound track is thereby eliminated. The lower magazine is pro- 
vided with a similar valve and porthole and is also equipped with a 
newly designed even-tension take-up. 

The improvements herein described have culminated a five-year 
period of designing and tooling-up, plus an exhaustive series of field 
tests in key circuit theaters operating fourteen hours daily over a 
span of sixteen months. 

A New Deluxe 35-Mm Motion 
Picture Projector Mechanism 





SUMMARY: Development of a new deluxe 35-mm motion picture projector 
mechanism, to be known as the RCA-100, was recently completed and produc- 
tion is now under way. The keynote in the design is high-quality projection 
continuously over a long period of time without the necessity of costly periodic 
factory overhauls and replacement of parts. Over ten years of field experi- 
ence with the Brenkert BX-80 projector mechanism has shown that auto- 
matic lubrication, a heavy rugged type of gear train, double rear shutters, unit 
construction and ease of serviceability are features essential to this objec- 
tive. These features, together with new additions necessary to meet pres- 
ent-day requirements will be described in this paper. 

Automatic Lubrication. In a well-lubricated system there is prac- 
tically no wear of the metal parts because all contact takes place on 
films of oil between mating surfaces. In this automatic lubrication 
system a geared pump inside the housing delivers a continuous flow 
of filtered oil through a copper tube from the oil reservoir in the 
base of the mechanism to a rotary lubricator at the top of the gear 
train. This rotary lubricator is perforated at longitudinal spacings 
so the various holes are in line with the plane of each gear and bearing 
in the gear train. In operation, oil is pumped from the reservoir to 
the rotary lubricator and then showered over all of the parts in the 
gear compartment, providing lubrication at the right places con- 

With this method of lubrication filtered oil is circulated throughout 
the entire gear side of the projector mechanism several times a minute. 
The heat generated in the intermittent is carried away by the cir- 
culating oil instead of remaining confined in the intermittent case. 
In this manner it also acts as an over-all cooling system in distributing 
local heat in the gear train throughout the whole mechanism. 

Figures 1 and 2 illustrate the advantages of the automatic system 
over the oilcan or pressure-feed method. All shafts and bearings 
in the gear train are designed so that they are lubricated continuously 

PRESENTED: April 27, 1950, at the SMPTE Convention in Chicago. 






Fig. 1. Pressure-feed and oilcan methods of lubrication. 

Fig. 2. Automatic lubrication. 

over their entire length without any oil leaking from the gear com- 
partment, as shown in Fig. 2. 

Gear Train. The gearing in the new projector, shown in Fig. 3, 
consists entirely of helical gears running on parallel shafts, except for 
the shutter-shaft drive gears which are spiral bevel gears. This 
type of gearing can be set up for minimum backlash and with the 
meshing teeth of mating gears contacting each other over the full 
width of the gear face. This means smooth and quiet operation and 

1950 DELUXE 35-Mn PROJECTOR 321 

negligible wear over a long period of time which, of course, means 
maintaining the original accuracy built into the mechanism. 

One of the other factors essential in correct gear design in a motion 
picture mechanism is to maintain a low gear ratio between the im- 
portant drive assemblies such as the intermittent, main-drive gear 
assembly and shutter-drive assembly. This keeps vibration at a low 
level, and prevents it from traveling through the mechanism where 
it could increase wear between gear teeth and allow the light shutters 

Fig. 3. Gear side of projector (gear cover removed). 

to oscillate. Excessive oscillation of the shutters results in inferior 
projection unless the shutters are widened to compensate for this 
movement, in which case the efficiency of light transmission would 

From a theoretical standpoint a gear ratio of 1 : 1 would be optimum 
for the least amount of wear, quietest operation and minimum vibra- 
tion. A 1 : 1 gear ratio is impossible throughout a gear train where 
shafts rotate at different relative speeds, however, so we have done 


the next best thing by using a 2 : 1 gear ratio between all important 
drive assemblies. 

The light-shutter compensator gear assembly has an important 
role in the gear train. The purpose of this unit is to enable the picture 
to be framed at the aperture while at the same time keeping the action 
of the light shutters in perfect time with the pull-down action of the 
intermittent, without the use of angular sliding gears. When the 
framing handle is turned, the radial positions of the shutter drive 
gears are changed with respect to each other but they always mesh 
with identically the same teeth in their mating gears, over the same 
portion of the gear face. With this type of shutter compensator it is 
not necessary to change the position of the framing knob periodically. 
The projector will run smoothly and quietly with the framing knob 
in any position; wear and backlash between gears are eliminated; 
the same good picture definition and high efficiency of light trans- 
mission to the screen obtained originally, is maintained throughout 
the life of the projector. 

Unit construction is used throughout the gear train. All assem- 
blies are accurately located by dowel pins so that perfect alignment 
is assured without fitting and adjusting for proper backlash. All 
units are completely interchangeable. 

Intermittent Mechanism. The intermittent, together with its star 
and cam, is shown in Fig. 4. All of the parts are large and heavily 
constructed. They can thus be manufactured with greater accuracy 
than could smaller ones and because of this the wear is negligible. 

A cross section of the star wheel, shaft, sprocket and bearings is 
shown in Fig. 5. The star wheel and sprocket shaft are supported 
for over 60% of the shaft's total length. This long bearing support 
and the extension of the bearings directly to the star wheel and to the 
sprocket cause these parts to be held with extreme precision. 

Bronze bearings are used throughout the intermittent because they 
can be manufactured with great accuracy and because of their long 
wearing qualities. Wear is reduced to a minimum through the use 
of this type of bearing with a continuous flow of oil through all parts 
of the intermittent unit. 

The index pin on the cam is fitted with a hardened steel roller to 
eliminate the possibility of flat spots developing on the index pin. 
This is another instance of precaution which was taken to reduce 
wear to a minimum. 

The intermittent can be removed and replaced in less time than is 
required to run a reel of film. The sprocket can be removed and 
replaced in less than one minute. 




Film Compartment. The film compartment as shown in Fig. 6 is 
enclosed by a large glass door with the visible interior illuminated by 
two concealed lights. This aids in accurate threading of the mech- 
anism and in easy inspection of all operating parts. 

The oil gage, which is an integral part of the oil pump, is located 

Fig. 4. Intermittent mechanism. 

6 5 
6 8 


"" ^ 



Fig. 5. Cross-section of star wheel, shaft and bearings. 



Large entrance and 

in the lower front corner of the main case, 
exit oil ports assure correct oil level indications. 

Unit construction is used throughout in the film compartment for 
accuracy and for easy servicing. 

Light Intercepting Shutters. The design of the light shutters and 
associated gearing in a motion picture projector mechanism determine 
the efficiency of light transmission to the screen. Consistently high, 

Fig. 6. Operating side of projector. 

efficient light transmission is of great importance in large indoor 
theaters and in drive-in theaters where pictures up to 70 ft in width 
are projected. Two rear shutters are used in the projector mech- 
anism rotating in opposite directions so that the light beam is cut 
simultaneously from the top and bottom. Wide experience has 
shown that double rear shutters are most desirable for the following 
reasons : 

1. The efficiency of light transmission is increased more than 20% 




above that which can be obtained from most projector mechanisms 
with a single shutter. 

2. The light beam is shadowed so that a black, cool aperture is 
obtained over 12% longer than when one front and one rear light 
shutter are used. 

3. Since the action of the light shutters is to cut the light beam 
from the top and bottom simultaneously, in a plane removed from 

Fig. 7. Complete projector and sound assembly. 

the film plane, the intensity of the light on the screen is gradually 
reduced to zero at the start of the pull-down and then gradually 
increased from zero to maximum at the end of the pull-down. High 
efficiency of light transmission can thus be obtained by designing the 
width of the shutter blades to take advantage of the small movement 
of film at the beginning and end of the pull-down period without any 
trace of a travel ghost. 


4. From both an operating and an appearance standpoint, it is 
more desirable for both shutters to be located at the rear of the mech- 
anism than to have one at the front and one at the rear. In those 
cases where a front shutter is used difficulty is sometimes experienced 
in removing the lens for cleaning, especially where the mechanism is 
located close to the front wall. 

Heat Baffle. The use of powerful arc lamps in many theaters today 
makes it essential to baffle all stray light from the metal parts of the 
film trap to prevent it from becoming excessively hot. A heat baffle 
is used which consists of three metal plates spaced about %Q in. apart 
and positioned in a vertical plane with the optical axis in such manner 
as to allow an //2.0 beam of light to be projected to the picture 
aperture with a minimum light spill around the metal parts of the 
film trap. 

The heat resulting from stray light intercepted by the heat baffle is 
carried away by a rotary fan located at the top of the main case, 
drawing air up past the sections of the baffle. 

The entire film trap is completely enclosed with a metal light shield 
preventing stray light escaping from the picture aperture and shining 
into the projectionist's eyes. 

Projection Lens Mount.- The lens mount has been designed to 
accommodate the new long focal length //2.0 projection lenses which 
are 4 in. in diameter. It is easily removed as a complete unit by the 
removal of four screws. A metal collar is provided with each pro- 
jector mechanism so that standard diameter projection lenses in focal 
lengths up to 5 in. can be accurately held in this lens mount. Two 
knurled thumb screws in split rings, one at each end of the lens mount, 
hold the projection lens rigidly and accurately in position. Focusing 
is done by means of a knob at the front of the lens mount which is 
accessible from either side of the mechanism. 

Figure 7 shows a complete projector assembly including the new 
Brenkert Supertensity Arc Lamp in combination with the new pro- 
jector. This is typical of the units which are now being supplied for 
deluxe drive-in theaters throughout the country. 

68th Convention 

RESERVATIONS are coming in to the Lake Placid Club and to the Hotel Marcy, 
these in response to the Convention Advance Notice which went to all members 
in mid-August. If you have overlooked yours, ask Society headquarters for the 
information and make your arrangements without delay. 

PAPERS have been scheduled for ten technical sessions; two evenings will be 

devoted to awards and Banquet; one evening is reserved for prerelease showing of 

a feature motion picture; and a prerelease feature motion picture will also be 

shown on one afternoon. Sessions topics, detailed in the Tentative Program 

being mailed separately, are : 

Monday Afternoon Television 

Monday Evening Award Presentation 

Tuesday Morning Television and TV Film Pictures 

Tuesday Afternoon Television, Sound Recording, Color 

Wednesday Morning Magnetic Recording 

Wednesday Afternoon High-speed Photography 

Wednesday Evening Cocktail Party, Banquet and Dance 

Thursday Morning High-Speed Photography 

Thursday Afternoon Film Registration, Aperture Calibration and Sound 


Thursday Evening Color and Trick Photography 

Friday Morning Sound, Projector Carbon and Theater Television 

Friday Afternoon Theater Television 

AT LAKE PLACID, there will be a program with many attractions: some new 
subjects and some generally familiar ones newly high-lighted; entertainment 
and recreation of inviting variety. 

Engineering Committees Activities 

Screen Brightness 

The Screen Brightness Committee under Wallace Lozier's Chairmanship is now 
ready to start the 100-theater screen brightness survey which has been under dis- 
cussion for the last six months. Actual measurements will begin about mid- 
September. Task groups responsible for survey work have been set up in Los 
Angeles, New York, Philadelphia, Chicago, Toledo and Rochester. The first 
theaters visited will be in the New York area, where it is planned to start with 30 
indoor and two outdoor theaters. 

The photoelectric instrument developed by Allen Stimson of the General Elec- 
tric Co. has been checked to assure accurate measurements, and since there is 
only one in existence the survey will necessarily have to proceed slowly at first. 
General Electric has agreed, however, to supply instruments for $345 each, pro- 
viding ten or more can be manufactured at one time. All likely customers are 
being canvassed, and it is hoped to have before long additional instruments avail- 
able to survey teams. 

Every means possible will be taken by those making the survey to avoid up- 
setting normal theater operation, and at least 24 hours' notice will be given any 
house it is proposed to survey. With the exception of about 15 minutes for mak- 
ing actual screen measurements, the remaining data can be gathered during the 
regular show. 


A word of thanks is due the International Projectionist for the excellent publicity 
they have given this project in both their June and July issues. We have every 
anticipation of a worth-while job being done. 


The first regular meeting of the Joint RTMA-SMPTE Committee on Television 
Film Equipment, was held at the Hotel New Yorker on July 18. Their work got 
off to an excellent start with all of the SMPTE delegates on hand. The primary 
task at the moment is the completion of a specification for a 16-mm television film 
projector which originated within RTMA. 

While the specification framework has been completed, many of the detail re- 
quirements need further study. Approximately a dozen task groups were or- 
ganized and requested to prepare drafts of various sections for circulation to com- 
mittee members prior to the next meeting. Standards for picture aperture size 
to be used in video recording and the area to be scanned in reproduction of opaques 
and slides were also discussed and recommendations will be made in the near 

Magnetic Recording 

Last April, Glenn Dimmick's subcommittee working on standards for magnetic 
recording recommended submitting proposed standards for track location on 35-, 
17y<r, 16- and 8-mm motion picture film to the Sound Committee for its recom- 
mendations on publication. The ballot was sent out early in July, but serious 
objections were received from one of the major studios which felt that the limited 
experience with the present proposals did not warrant wide circulation in the 
JOURNAL. Further action will be delayed until this problem is resolved within the 
Sound Committee. 

High-Speed Photography Question Box 

Here are answers to five questions on 
high-speed photographic techniques 
which appeared on p. 122 of the July 
JOURNAL. These answers were contrib- 
uted by: J. H. Waddell of Wollensak 
Optical Co.; Henry M. Lester, Consult- 
ant; Kenneth Shaftan of Burke and 
James; and Eugene L. Perrine of the 
Armour Research Foundation. 

Further questions and answers will 
appear in subsequent JOURNALS. If you 
wish to participate send either your 
questions or answers to Society Head- 

A 1 This question concerned tak- 
** * ing high-speed motion pictures 
of moving parts inside a black bakelite 
device the size of a dime. Speeds of 
4,000 to 8,000 frames/sec were required. 
With methods now being used, insuf- 
ficient exposure has been obtained when 
using Super XX film and heat generated 


by the light source altered performance 
of the device under test. 

One suggested solution was the use of 
continuous flash lighting units to pro- 
.vide ample light free of heating effects 
normally encountered with tungsten or 
arc illumination. Adequate exposure 
and depth of field can be secured by 
using two flash units properly placed, 
and a 2-in. lens with a suitable extension 
tube at effective apertures ranging from 


Figure A-l. 

A second solution is proposed in Figure 
A-l. In this method the center was cut 

out of a 12-in. diameter parabolic mirror 
of 6-in. focal length so that it could slip 
over the lens and extension tube of the 
camera. Precision quality Bausch & 
Lomb parabolic mirror has been used. 
Elliptical mirrors arc better suited for 
this application when photographing 
extremely small areas, but where sharp 
focus of the light beam is not required 
the paraboloid is satisfactory. A small, 
hand-fed carbon arc using from 5 to 10 
amp is used about 2 ft behind the sub- 
ject, and the mirror adjusted so that the 
arc is focused on the area to be photo- 
graphed. A water cell for removing the 
heat is placed in the beam close to the 
arc so that all light falling on the mirror 
passes through the cell. The proponent 
of this method states that sufficient 
illumination is obtained for magnifica- 
tion on the film of up to five times at a 
speed of 5000 frames/sec. 

A third proposed solution uses a plane 
mirror with an elliptical hole large 
enough to accept the camera lens. The 
mirror is placed at an angle of 45 to 
the optical axis so that the light is re- 
flected from the mirror to the subject. 
A partial reflection transmission mirror 
could be used instead, but that would 
reduce the exposure by a factor of ap- 
proximately 50%, and only half the 
light would reach the subject. Satis- 
factory mirrors of this type may be 
secured from Evaporated Films, Inc., 
Ithaca, N.Y. It is also recommended 
that long focus lenses with extension 
tubes be used to secure adequate dis- 
tance between the subject and the 
camera. If a water cell is used, as de- 
scribed on p. 450 in the article "High- 
Speed Photography" in the November 
1949 JOURNAL, heat can be reduced to a 
negligible amount. It is also suggested 
that a G.E. 750-R lamp or a Rosslite 
of similar characteristics be employed. 
Distance from lamp to mirror to subject 
should be approximately 15 in. for 
maximum illumination. In making 
high-speed pictures of this type, a suit- 
able exposure meter should always be 

A O The second question con- 
** * cerned high-speed motion pic- 
tures of small parts of a mechanical de- 
vice moving at 15 to 30 cycles/sec. A 
Fastax camera is employed at a frame 
rate of 1250 frames/sec, with a 6-in. 

lens, an object distance of 8 ft, Super 
XX reversal film, and two 750-w re- 
flector spot lamps. Since all surfaces 
had similar finish, it was extremely dif- 
ficult to distinguish between adjacent 
parts in the projected picture. 

The first reply to this question sug- 
gested very diffuse lighting through use 
of a translucent tent between light and 
subject. It was pointed out that this 
would obviously result in a considerable 
loss of light, but with continuous flash 
lighting units this loss could be toler- 
ated. By appropriate arrangement of 
the tent and choice of material, however, 
loss of light can be held to a minimum. 

A second answer stated that if the 
light source is placed correctly, there 
should not be too much trouble from 
specular reflections when the exposure 
factor is correct. Bad flare is produced 
from machined parts when there is defi- 
nite over-exposure in a high speed camera. 
If the exposure is somewhat reduced, 
brightness of parts, even though made of 
brightly polished metal, should be easily 
controlled. The light source must be 
as near the camera as possible, and either 
G.E. Electric 750-R or Rosslite lamps 
should be used. 

A O Question 3 dealt with photo- 
* ** graphing vibration effect on 
various components of air-borne instru- 
ments. These instruments are extremely 
small and encased, making it necessary 
to illuminate and photograph through a 
hole in the cover. Vibration frequencies 
of 800 cycles/sec, with object motion as 
little as 0.001 in. are encountered. 

The first reply pointed out that it is 
possible to photograph and illuminate 
through a hole in the cover of an encased 
instrument by high-speed photography 
only if the hole is large enough. The 
smallest hole believed to be feasible is 
about 5 in. in diameter. It was stated, 
however, that a somewhat smaller hole 
might be used with variations in tech- 

The first method suggested was to 
direct the light output of a continuous 
flash lighting unit on a spherical mirror 
with a hole in the center for the camera 
lens. A second method was to surround 
the camera lens with an electronic flash 
tube, discharging its light output in 
synchronism with the high-speed camera 
shutter. It was pointed out that Dr. 


Harold Edgerton of M.I.T. has de- 
signed an electronic flash lamp capable 
of doing this job. For conditions out- 
lined in this question, it was suggested 
an Eastman Type 3 camera be used at 
frame rates of 3000 frames/sec. A 
movement of 0.001 in. could then be 
magnified about 200 times both in time 
and space, offering an adequate record, 
either on the screen or in still picture 
enlargements of single frames. 

Another reply suggested use of a Fas- 
tax camera with auxiliary control equip- 
ment to secure 14,000 pictures/sec. 
Frame rates of this order are necessary 
for studying frequencies as high as 800 
cycles/sec. It was also suggested that 
in studying vibrations of extremely small 
excursion extension tubes be used on 
camera lenses, and the pictures be pro- 
jected at about a magnification of 100. 
Magnification of 100 times of a 0.001- 
in. excursion will then appear on the 
screen as 0.1 in. For lighting, a G.E. 
750-R lamp should be used, so placed 
that the plane of vibration is clearly 
emphasized with respect to the station- 
ary surrounding subject. At least two 
lamps should be used in this setup with 
a high-low series-parallel switch in order 
to focus the camera with lamps in series 
and expose with lamps in parallel. 

A A This question dealt with 
** " photographing a 3 X 5 ft 
area of a dark machine at a frame rate of 
3000/sec. Here again inadequate ex- 
posure was being obtained, and high 
amperage power lines were not available. 
The first reply stated that successful 
results had been obtained under similar 
conditions, using continuous flash light- 
ing units on dark areas of up to 4 sq ft 
In this case, also, the machine being 
photographed was black. The frame 
rate was 3000/sec, with the lens set at 

//4. Using Super XX film, a satisfac- 
tory record was attained. This type of 
lighting requires much less power than 
incandescent units. 

A second reply suggests use of sunlight 
for illumination. If the equipment pho- 
tographed is extremely dark, it may be 
necessary to use booster mirrors to 
light adequately the whole surface. 
Frame rates of 3000/sec are entirely 
possible in direct sunlight, but not be- 
hind windows. 

A C This question dealt with spe- 
*"* *J cial processing for reversal 
film used in high-speed photography. 
The first reply named two manufactur- 
ers of processing equipment suitable 
for this type of work: Micro Record 
Corp., New York City; and Morse 
Instrument Co., Hudson, Ohio. It 
was pointed out, however, that while 
machines made by either of these com- 
panies could do a job of controlled proc- 
essing, in both cases the task is tedious 
and far from satisfactory when a quan- 
tity of film is involved. Both require 
special drying facilities and great care 
in handling of films with black coatings. 
It was believed that the advantages of 
longer first development are questionable 
unless the additional development is 
accurately timed and definitely related 
to the degree of underexposure. Faster 
film such as Kodak's Linagraph (nega- 
tive stock is 50% to 60% faster than 
Super XX reversal) might be used, and 
is simpler to process on the two units 
mentioned above. The best answer is to 
avoid working near the borderline of 
underexposure, which always results in 
pictures lacking in detail, definition, 
contrast and depth. 

Another reply suggested referring 
this problem to the Houston Corp. in 
Los Angeles which builds special 16-mm 
processing equipment. 

Journals Out of Stock: The Society's stock of JOURNAL issues for March, 
Part II, July, August and September, 1949, has been exhausted as a 
result of an unexpected increase in demand and the Society's Head- 
quarters is anxious to purchase a stock of each. Members or libraries 
having extra copies available are invited to send them in. The going 
price is 75c. 


Book Reviews 

The American Annual of Photography, Volume 64, 1950. Edited 
by Frank R. Fraprie and Franklin I. Jordan 

Published (1949) by the American Photographic Publishing Company, St. 
Paul, Minn. 208 pp. incl. 120 illus. + 38 pp. Who's Who + 30 pp. Advt. Paper 
bound, 7 X 10 in. Price, $2.00. 

This 1950 issue (Volume 64), I feel, surpasses all previous issues. Articles such 
as "The Work of Jose Ortiz-Echagiie" are entertaining and inspiring, especially 
when so splendidly illustrated. Other articles are equally well written and illus- 
trated. Such articles as "Printing Exposure Determination by Photoelectric 
Methods" and "The Physiology of Film Base" will probably appeal more to our 
technically minded SMPTE members but such articles as "The Motion Picture 
Camera in Science and Industry," "The Camera as a Field Research Tool," 
"Photography in Industry and Science," "The Work of Eadweard Muybridge," in 
fact all of the sixteen diversified articles will appeal to anyone interested in the 
progress of photography, pictorial and otherwise. 

There are some 67 full-page pictorial illustrations of an international nature 
which are intelligently described and analyzed by Frank R. Fraprie. 

The Who's Who in Pictorial and Color Photography as well as the exhibition 
records for the past three years will be of special interest to those who are con- 
cerned with salon exhibition. JOHN W. BOYLE, 139V2 S. Doheny Drive, Los 
Angeles 48, Calif. 

Practical Television Engineering, by Scott Helt 

Published (1950) by Murray Hill Books, Inc. (A subsidiary of Rinehart and 
Co.), 232 Madison Ave., New York 16, N.Y. xv, 708 pp., including 30 pp. glos- 
sary, 14 pp. index, 387 illus. and numerous tables. 6 X 9 in. Price, $7.50. 

Mr. Helt's book is a significant contribution to the television-engineer-to-be. 
With the lifting of the freeze, the rush to install more television stations will be on 
in full force. Many electronics engineers will be faced for the first time with the 
day-to-day television operating problems. It appears that Mr. Helt was aiming 
toward that group particularly. They will find this book extremely helpful. 

There is a certain unevenness in the density of theoretical treatment. Upon 
analysis, it becomes evident that this is just what Mr. Helt intended. For ex- 
ample, the section on studio lighting is right to the point with details of the type of 
lights to use and how to place them. Yet the theory of the image orthicon is cov- 
ered in simple straight-forward language minus equations. This makes good 
sense because no operating engineer is going to design an image orthicon. He has 
only to recognize its operating characteristics and decide whether or not a tube 
should be used or rejected. Yet with lighting he can be a "designer" and with 
this book he has sufficient information to deal intelligently with the problem with- 
out reference to any other source material. 

The discussion of lens theory is well handled and the bridge to electron optics 
skillfully presented. The advanced reader is naturally led to more rigorous texts 
on electron optics. 

The importance of the cathode-ray tube oscilloscope to the television engineer 
cannot be overemphasized. Mr. Helt wisely goes into great detail to explain its 
operation and use. This chapter alone will make this book very important. He 
also gives interesting manufacturing information on cathode-ray tubes which 
provides the new television engineer with some usefu^ background. 


The chapter devoted to the synchronizing generator is quite complete. The 
theory and design concepts are well presented, particularly where they will pro- 
vide a better understanding necessary to good maintenance technique. The 
succeeding chapters deal competently with video amplifiers and associated com- 
pensating circuits, power supplies and the receiver. 

Mr. Helt makes a successful effort to provide the operating television broadcast 
engineer with a good understanding of the receiver. Too often, engineers over- 
specialize to a point where the station man has little understanding of the receiver 
man's problem. Yet no television system is complete without the home receiver. 

With regard to the transmitter, more detailed information may be required. 
The author favors the studio engineer by providing helpful hints on approved 
maintenance procedures and best studio practice. The book is already long, 
nearly 700 pages. 

Mr. Helt has succeeded in authoring a book which was greatly needed. He has 
accomplished his task with a professional quality. This book is fully recom- 
mended to the industry as a practical exposition of the engineering problems in 
television broadcasting. E. ARTHUR HUNGERFORD, JR., General Precision Labora- 
tory, Pleasantville, N.Y. 

Sound Absorbing Materials, by C. Zwikker and C. W. Kosten 

Published (1949) by Elsevier Publishing Co., 215 Fourth Ave., New York 3. 
171 pages + 3 pp. index. 92 illus. 9% X 63^ in. Price, $3.00. 

The first-named author was formerly Professor of Physics at Delft Technical 
University, Netherlands, and is now connected with Philips Electrical Indus- 
tries, Eindhoven. The second is Lecturer of Physics at Delft Technical Univer- 
sity. Their book is essentially an account of the theoretical and experimental 
work done by them and by other European investigators, with some references to 
American sources, in developing along basic scientific lines the relation of the 
sound absorbing properties of materials to measureable physical characteristics of 
their composition and structure. The first chapter treats the use of acoustic 
impedance as a valuable intermediate step in this relation. In later sections the 
wave equations and impedance characteristics are derived for several types of 
absorbing media: an air-impervious compressible material with internal friction 
(sponge rubber), a porous material with an elastic frame (felt or mineral wool 
blanket), and a porous material with a relatively rigid frame, as exemplified by 
some of the common types of commercial acoustical materials. Methods of 
measurement of the material constants governing impedance, such as air-flow 
resistance, porosity (percent of voids), and compression modulus of the material 
structure are discussed. Measurement of impedance and normal incidence coeffi- 
cients of small samples is covered in some detail, and typical experimental results 
are given. 

Absorption by resonators is treated extensively. These include the simple 
Helmholtz resonator consisting of an air cavity with a small orifice and combina- 
tions of such resonators having staggered frequency responses. Practical con- 
structions of this type have been used successfully in Europe for room acoustical 
correction. The basic resonator theory is extended to the more familiar case of a 
perforated rigid board over an air space which may be completely or partially 
filled with porous absorbing material. Useful design formulae and charts are 
included for the various cases. It is rather surprising that no mention is made of 
absorption by diaphragmatic vibration, which is utilized in the familiar curved 
plywood studio treatments and in at least one commercial material. Another 


distinct type of absorber is the integrally perforated porous material. This is 
very widely used, but is touched on only briefly in the book, and no attempt is 
made to develop an adequate theory for this case. 

In the final chapter, methods of absorption measurement at angles of incidence 
other than normal are briefly mentioned, and the difficulties in predicting ab- 
sorption characteristics under random incidence or room conditions from normal 
incidence data are pointed out. HALE J. SABINE, The Cclotex Corp., Chicago 3, 111 . 

American Cinematographer Hand Book and Reference Guide 

Seventh Edition, by Jackson J. Rose 

Published (1950) by American Cinematographer Hand Book, 1165 North 
Berendo St., Hollywood 27, Calif. 299 pp., 3 pp. index, 85 tables, 10 photo- 
graphs in color + 38 pp. advt. GVa X 4 in. Flexible binding. Price, $5.00. 

The Seventh Edition of this convenient pocket size hand book and reference 
guide has been expanded to 325 pages. It still contains the charts, formulas and 
technical information which professional cinematographers have been using for 
years but the book has been brought up to date with the addition of latest informa- 
tion on the various color processes : Technicolor, Monopack, Ansco, Kodachrome, 
Du Pont, Ektachrome, Bipack, Trucolor, etc. The new method of "Intensifica- 
tion" is explained as well as many of the newer gadgets being used in the profes- 
sional field today. The color illustrations are extremely helpful in showing vari- 
ous "filter" results in monochrome. Magnetic recording, television photography 
and "T" stops are a few of the newer subjects. The author and compiler, Jackson 
J. Rose, A.S.C., has had the cooperation of his colleagues in the film industry and 
has been quick to use their suggestions for improving cinematography and finding 
a simpler way to achieve artistic photographic results. JOHN W. BOYLE, 139V2 S. 
Doheny Drive, Los Angeles, 48, Calif. 

Theatre Catalog, 8th Annual Edition, 1949-1950 

Published (1950) by Jay Emanuel Publications, Inc., 1225 Vine St., Philadel- 
phia 7. 1-528 pp. 4- i-x, profusely illus., includes advtg. 9M X 12^t in. Price 
$5.00 (foreign shipments $10.00 a copy). 

This new Theatre Catalogue isn't the type of publication that motion picture and 
television engineers would normally read. It is, nevertheless, an impartial 
picture of motion picture operation and design, covering almost every phase of a 
fascinating business. 

The engineer's interest in this great industry cannot properly be limited to his 
laboratory. Auditorium design is changing constantly and with it new problems 
confront the alerted engineer. Panoramic viewing conditions approaching the 
normal viewing conditions of the human eye are desirable, yet little has been 
done about it. Drive-in theaters are here to stay and so is theater television. 
Third dimension projection is a stimulus that theaters need badly. What is 
being done about it today? 

The Theatre Catalogue not only discusses certain phases of projection and 
sound but dwells on design and construction, maintenance and management. 
The engineer must be familiar with these phases of the business, otherwise he 
cannot properly tackle theater operation problems. 


Attention is directed, for instance, to the section on theater design and to the 
section on new equipment. Know well the ultimate use of equipment so care- 
fully designed in the laboratory. Where and under what conditions will the 
finished motion picture be viewed by John Public? How can improvements be 
made in the over-all result? How can picture presentation be vitalized? What 
changes can be made to better a system of projection now essentially 23 years old? 

The Theatre Catalogue is not primarily reading matter for an engineer but it 
should be. By completely understanding a vast operation it is hoped that the 
motion picture and television engineers will: (1) see the inadequacy of current 
practices so that they may be improved; (2) publish the results of their findings 
freely so that others may develop the germ of an idea; (3) realize that they are 
likely to play as important a part as anyone else in this business' future; and (4) 
believe that their ideas are good as far as they go but that they do not go far 
enough. LEONARD SATZ, Ray tone Screen Corp., 165 Clermont Ave., Brooklyn, 

Current Literature 

THE EDITORS present for convenient reference a list of articles dealing with sub- 
jects cognate to motion picture engineering published in a number of selected 
journals. Photostatic or microfilm copies of articles in magazines that are avail- 
able may be obtained from The Library of Congress, Washington, D.C., or from 
the New York Public Library, New York, N.Y., at prevailing rates. 

American Cinematographer 

vol. 31, no. 5, May 1950 
Pushbutton Zoom Lens for TV 

(p. 160) H. I. SMITH 
Adapting Motion Picture Lighting to 

Television (p. 162) L. ALLEN 

vol. 31, no. 6, June 1950 
The Infra-Red Photographer Evaluator 

(p. 196) S. HORSLEY 
Matching Location Footage with 

Studio Shots (p. 197) H. A. LIGHT- 
Optical Effects with Any Camera 

(p. 198) I. BROWNING 
When and How to Use Camera Angles 

(p. 201) P. TANNURA 
Britons First With Tape Sound Unit 

for Silent Home Movie Projector 

(p. 204) 
Ansco Announces New 16-mm Color 

Duplicating Film (p. 205) 

Audio Engineering 

vol. 34, no. 6, June 1950 

The Columbia Hot Stylus Recording 
Technique (p. 11) W..S. BACHMAN 

An Adventure hi Loudspeaker Design 
(p. 14) H. T. SOUTHER 

Considerations in the Design of Feed- 
back Amplifiers (p. 17) H. I. KEROES 

International Photographer 

vol. 22, no. 5, May 1950 
Are Cameramen Necessary on TV? 
(p. 5) H. BIRCH 


The Camera Optical Engineer (p. 8) 

International Projectionist 

vol. 25, no. 5, May 1950 
Notes on Modern Projector Design 
(p. 14) R. A. MITCHELL 

vol. 25, no. 6, June 1950 
Notes on Modern Projector Design, 

Pt. II (p. 7) R. A. MITCHELL 
Heat, Light Reflectivity is Upped by 

Kodak Mirror (p. 11) 
An Optical Alignment Check System 

(p. 17) C. W. HANDLEY 
New Simplex Sound System Shown by 

IPC (p. 23) 
U. S. Navy 16-mm Projection Specs 

(p. 26) J. J. McCORMICK 

Motion Picture Herald 

vol. 180, no. 1, July 1, 1950 
Safety Stock is Now 85% in Use by 
Trade (p. 13) 

Radio & Television News 

vol. 43, no. 6, June 1950 
RCA's New Direct-view Tri-color 
Kinescopes (p. 46) 


vol. 9, no. 7, July 1950 
Experimental Tri-Color Cathode Rav 

Tube (p. 34) C. S. SZEGHO 
Process Screen Projection, Pt. I 

(p. 39) R. A. LYNN and E. P. 


New Members 

The following have been added to the Society's rolls since the list published last month. 
The designations of grades are the same as those in the 1950 MEMBERSHIP DIRECTORY: 
Honorary (H) Fellow (F) Active (M) Associate (A) Student (S) 

Baumhofer, Hermine M., Archivist, 
Wright-Patterson Air Force Base. 
Mail: 532 Telford Ave., Dayton 9, 
Ohio. (A) 

Cleveland, H. W., Physicist, Eastman 
Kodak Co. Mail: 1669 Lake Ave., 
Rochester, N.Y. (A) 

Crose, Harold G., Motion Picture Pro- 
jectionist, Lyric & Rialto Theatres. 
Mail: 855 S. 20th East St., Salt Lake 
City, Utah. (A) 

Del Porte, Earle N., Projection Super- 
visor, Station KSD-TV. Mail: 445 
Alice Ave., Kirkwood 22, Mo. (A) 

Downes, L. C., Designer, TV Film Pro- 
jection Equipment, General Electric 
Co. Mail: 947 James St., Syracuse, 
N.Y. (A) 

Duggan, Robert, Owner, The Studio 
Lighting Co., 1548 N. Dearborn, 
Chicago, 111. (A) 

Fallen, Louis F., Sales Representative, 
Ampro Corp. Mail: 985 Franklin 
Turnpike, Allendale, N. J. (A) 

Fulgham, Claude O., Vice-President in 
Charge of Management, Video Thea- 
tres. Mai ] IP/2 N. Lee, Box 1334, 
Oklahoma City, Okla. (A) 

Julin, Kurt, Technical Chief, A. B. Cos- 
morama. Mail: Skillnadsgatan 60A, 
Gothenburg, Sweden. (M) 

Kinstler, Richard C., Head, Photographic 
Laboratory. Procter & Gamble Co., 
M. A. & R. Bldg., Cincinnati 17, Ohio. 

Lepore, Alfred Louis, Electro-Acoustic 
Engineer and Cameraman. Mail: 732 
and 736 Man ton Ave., Providence 9, 
R.I. (M) 

Nemeth, Ted, Motion Picture Producer, 
Director and Cameraman, Ted Nemeth 
Studios, 729 Seventh Ave., New York 
19. (M) 

Parker, Will A., Motion Picture and 

Television Consultant, Film Counselors, 
Inc. Mail: 60 Manursing Ave., Rye, 
N.Y. (A) 

Potts, Clifford F., Motion Picture Pro- 
ducer, Fordel Film Laboratories, 1187 
University Ave., Bronx 52, N.Y. (M) 

Rivera, Joseph V., General Motion Pic- 
ture Laboratory work and Dupe Print- 
er, Circle Film Laboratory. Mail: 
873 E. 162 St., Bronx 59, N.Y. (A) 

Saunders, James Arthur, Assistant Engi- 
neer, Western Australian Government. 
Mail: 257 Crawford Rd., Ingle wood, 
Western Australia. (A) 

Shagin, Ralph J., Photographic Mer- 
chandising Analyst. Mail: 686 Kent 
Ave., Teaneck, N.J. (M) 

Temple, Dwight Irving, Television Engi- 
neer, Technical Supervisor, Columbia 
Broadcasting System, Inc. Mail: 47 
Lockwood Ave., New Rochelle, N.Y. 

Watson, Lloyd E., Research Chemist, 
Technicolor Motion Picture Corp. 
Mail: 1708 Scott Rd., Burbank, Calif. 

Willoughby, Anthony Haydn, Consultant 
Electrical Engineer, Sir Robert Wat- 
son-Watt & Partners, Ltd. Mail: 7 
Gay fere St., Westminster, London, 
England. (A) 


Anderson, James A., Assistant Produc- 
tion Manager, Alexander Film Co., 
Colorado Springs, Colo. (A) to (M) 

Churko, John G., Sales Engineer, Century 
Lighting Co., Inc. Mail: 106 E. 108 
St., New York 29, N. Y. (S) to (A) 

Williams, Paul A., Audio-Video Engineer, 
KPIX, Inc. Mail: 341 Hazelwood 
Ave., San Francisco 12, Calif. (A) to 

SMPTE Officers and Committees: The roster of Society Officers 
was published in the May JOURNAL. The Committee Chairmen and 
Members were shown in the April JOURNAL, pp. 515-22; changes in 
the Engineers Committees have been extensive and so the complete 
rosters are given on pp. 337-40 of this JOURNAL. 


New Products 

Further information concerning the material described below can be 
obtained by writing direct to the manufacturers. As in the case of 
technical papers, publication of these news items does not constitute 
endorsement of the manufacturer's statements nor of his products. 



The Spectra Three-Color Meter is a new instrument recently announced by 
Photo Research Corp., Burbank, Calif., succeeding the firm's Spectra color 
temperature meter widely used in motion picture color photography. 

Designers of the new instrument have spent several years developing a system 
which would relate the amounts of red, green and blue in an illuminant to the 
color balance of different types of color film and to the selection of any necessary 
corrective filters. The result is a log index derived from the ratios of blue to red 
and of green to red. The Spectra Index for photoflood lamps is 2.0/1.0 which 
means that a photoflood lamp emitting light of the color for which Type A color 
film is balanced will give a reading of 2.0 on the blue-red scale of the meter, and of 
1.0 on the green-red scale. 

If the B-R reading is more than one unit high or low, a correction filter of the 
turquoise-salmon series must be applied. If the G-R reading is more than half a 
unit away from the correct value, a correction filter of the green-magenta series 
must be placed over the lens. In either case, a computer (right, above) indicates 
directly the filter to employ. A card is furnished indicating the Spectra Index 
ratings of available color film materials. 

To facilitate the use of the three-color system, the firm is also making a complete 
series of mounted glass filters to match the scales of the Three-Color Spectra. 
One series of filters, the C7 1 , provide the usual type of correction for yellowish or 
bluish light. The new series, the GC, correct for a deficiency or excess of green 
in the illuminant. 

In addition to supplying the new meter, Photo Research Corp. reports that it 
will convert any of the older two-color meters to the new model, at a reasonable 
charge, the shape and general construction of the instrument having been kept the 


Society Engineering Committees 

As OF AUGUST 15, 1950 

CINEMATOGRAPHY. To make recommendations and prepare specifications for the opera- 
tion, maintenance, and servicing of motion picture cameras, accessory equipment, studio and 
outdoor-set lighting arrangements, camera technique and the varied uses of motion picture 
negative films for general photography. (File 5) 

C. G. Clarke, Chairman, 20th Century-Fox Film Corp., Beverly Hills, Calif. 

( Under organization) 

COLOR. To make recommendations and prepare specifications for the operation, maintenance, 
and servicing of color motion picture processes, accessory equipment, studio lighting, selection 
of studio set colors, color cameras, color motion picture films, and general color photography. 
(File 10) 

H. H. Duerr, Chairman, Ansco, Binghamton, N.Y. 

R. H. Bingham R. O. Drew L. T. Goldsmith C. F. J. Overhage 

M. R. Boyer A. A. Duryea A. M. Gundelfinger W. E. Pohl 

H. E. Bragg R. M. Evans W. W. Lozier G. F. Rackett 

O. O. Ceccarini J. G. Frayne A. J. Miller L. E. Varden 

FILM DIMENSIONS. To make recommendations and prepare specifications on those film 
dimensions which affect performance and inter changeability, and to investigate new methods 
of cutting and perforating motion picture film in addition to the study of its physical properties. 
(File 15) 

E. K. Carver, Chairman, Eastman Kodak Co., Kodak Park Works, Rochester 4, N.Y. 
E. A. Bertram W. G. Hill N. L. Simmons Fred Waller 

A. F. Edouart A. J. Miller M. G. Townsley D. R. White 

A. M. Gundelfinger W. E. Pohl William Wade 

FILM-PROJECTION PRACTICE. To make recommendations and prepare specifications 
for the operation, maintenance, and servicing of motion picture projection equipment, projec- 
tion rooms, film-storage facilities, stage arrangement, screen dimensions and placement, and 
maintenance of loudspeakers to improve the quality of reproduced sound and the quality of the 
projected picture in the theater. (File 20} 

L. W. Davee, Chairman, Century Projector Corp., 729 Seventh Ave., New York 19 
C. S. Ashcraft C. F. Horstman Paul Ries Ben Schlanger 

Frank Cahill G. T. Lorance Harry Rubin J. W. Servies 

R. H. Heacock H. T. Matthews 

HIGH-SPEED PHOTOGRAPHY. To make recommendations and prepare specifications 
for the construction, installation, operation, and servicing of equipment for photographing and 
projecting pictures taken at high repetition rates or with extremely short exposure times. 
(File 25) 

J. H. Waddell, Chairman. Wollensak Optical Co., 850 Hudson Ave., Rochester 5, N.Y. 

H. E. Edgerton, Vice-Chairman, Dept. of Electrical Engineering, Massachusetts Institute of 

Technology, Cambridge 39, Mass. 

E. A. Andres, Sr. W. R. Fraser C. D. Miller Earl Quinn 

K. M. Baird W. H. Fritz A. P. Neyhart M. L. Sandell 

D. M. Beard Eleanor Gerlach W. S. Nivison Kenneth Shaftan 

H. W. Crouch C. C. Herring Brian O'Brien C. W. Wyckoff 

C. H. Elmer H. M. Lester D. H. Peterson A. M. Zarem 

R. E. Farnham L. R. Martin 


LABORATORY PRACTICE. To make recommendations and prepare specifications for the 
operation, maintenance, and servicing of motion picture printers, processing machines, in- 
spection projectors, splicing machines, film-cleaning and treating equipment, rewinding 
equipment, any type of film-handling accessories, methods, and processes which offer in- 
creased efficiency and improvements in the photographic quality of the final print. (File 30) 

J. G. Stott, Chairman, Du Art Film Laboratories, 245 West 55 St., New York, N.Y. 
V. D. Armstrong I. M. Ewig O. W. Murray V. C. Shaner 

H. L. Baumbach T. M. Ingman W. F. Offenhauser, J. H. Spray 

D. P. Boyle P. A. Kaufman Jr. Lloyd Thompson 

O. E. Cantor C. F. LoBalbo W. E. Pohl Paul Zeff 

Gordon Chambers J. A. Maurer E. H. Reichard 

MOTION PICTURE STUDIO LIGHTING. To make recommendations and prepare specifi- 
cations for the operation, maintenance, and servicing of all types of studio and outdoor auxil- 
iary lighting equipment, tungsten light and carbon-arc sources, lighting-effect devices, diffusers, 
special light screens, etc., to increase the general engineering knowledge of the art. (File 35} 

M. A. Hankins, Chairman, Mole-Richardson Co., 937 N. Sycamore Ave., Hollywood 38, Calif. 
Richard Blount Karl Freund C. R. Long D. W. Prideaux 

J. W. Boyle C. W. Handley W. W. Lozier Petro Vlahos 

OPTICS. To make recommendations and prepare specifications on all subjects connected with 
lenses and their properties. (File 40) 

R. Kingslake, Chairman, Eastman Kodak Co., Hawk Eye Works, Rochester 4, N.Y. 
F. G. Back J. W. Gillon G. A. Mitchell L. T. Sachtleben 

A. A. Cook Grover Laube A. E. Murray O. H. Schade 

C. R. Daily J. A. Maurer W. E. Pohl M. G. Townsley 

I. C. Gardner 

PRESERVATION OF FILM. To make recommendations and prepare specifications on methods 
of treating and storage of motion picture film for active, archival, and permanent record pur- 
poses, so far as can be prepared within both the economic and historical value of the films. 
(File 46) 

J. W. Cummings, Chairman, National Archives, Washington 25, D.C. 

Henry Anderson C. R. Fordyce A. C. Hutton W. E. Pohl 

W. G. Brennan J. E. Gibson J. B. McCullough W. D. Stump 

J. W. Dunham G. Graham N. F. Oakley 

PROCESS PHOTOGRAPHY. To make recommendations and prepare specifications on motion 
picture optical printers, process projectors (background process), matte processes, special 
process lighting technique, special processing machines, miniature-set requirements, special- 
effects devices, and the like, that will lead to improvement in this phase of the production art. 
(File 50) 

M. H. Chamberlin, Chairman, Metro-Goldwyn-Mayer Studios, Culver City, Calif. 

(Under Organization) 

SCREEN BRIGHTNESS. To make recommendations, prepare specifications, and test methods 
for determining and standardizing the brightness of the motion picture screen image at various 
parts of the screen, and for special means or devices in the projection room adapted to the 
control or improvement of screen brightness. (File 55) 

W. W. Lozier, Chairman, National Carbon Div., Fostoria, Ohio 

Herbert Barnett L. D. Grignon L. J. Patton C. R. Underbill, Jr. 

F. E. Carlson A. J. Hatch, Jr. J. W. Servies H. E. White 

Gordon Edwards L. B. Isaac B. A. Silard A. T. Williams 

E. R. Geib F. J. Kolb Allen Stimson D. L. Williams 

L. T. Goldsmith W. F. Little 


16- MM AND 8-MM MOTION PICTURES. To make recommendations and prepare specifi- 
cations for 16-mm and 8-mm cameras, 16-mm sound recorders and sound-recording practices, 
16-mm and 8-mm printers and other film laboratory equipment and practices, 16-mm and 8- 
mm projectors, splicing machines, screen dimensions and placement, loudspeaker output 
and placement, preview or theater arrangements, test films, and the like, which will improve 
the quality of 16-mm and 8-mm motion pictures. (File 60) 

H. J. Hood, Chairman, Eastman Kodak Co., 343 State St., Rochester 4, N.Y. 

H. W. Bauman G. A. Del Valle D. F. Lyman A. G. Petrasek 

W. C. Bowen J. W. Evans W. C. Miller A. C. Robertson 

F. L. Brethauer C. R. Fordyce J. R. Montgomery L. T. Sachtleben 

F. E. Brooker John Forrest J. W. Moore H. H. Strong 

F. E. Carlson R. C. Holslag W. H. Offenhauser, Lloyd Thompson 
S. L. Chertok Rudolf Kingslake Jr. M. G. Townsley 

E. W. D'Arcy W. W. Lozier 

SOUND. To make recommendations and prepare specifications for the operation, maintenance, 
and servicing of motion picture film, sound recorders, re-recorders, and reproducing equip- 
ment, methods of recording sound, sound- film processing, and the like, to obtain means of 
standardizing procedures that will result in the production of better uniform quality sound in 
the theater. (File 66} 

L. T. Goldsmith, Chairman, Eastman Kodak Co., 343 State St., Rochester 4, N.Y. 

G. L. Dimmick, Vice- Chairman, RCA Victor Division, Camden, N.J. 

F. G. Albin R. M. Fraser E. W. Kellogg G. E. Sawyer 
A. C. Blaney J. G. Frayne J. P. Livadary R. R. Scoville 

D. J. Bloomberg L. D. Grignon K. M. Macllvain W. L. Thayer 

F . E. Cahill, Jr. Robert Herr W. C. Miller M. G. Townsley 

E. W. D'Arcy J. K. Hilliard G. C. Misener R. T. Van Niman 
R. J. Engler L. B. Isaac Otto Sandvik D. R. White 

STANDARDS. To survey constantly all engineering phases of motion picture production, dis- 
tribution, and exhibition, to make recommendations and prepare specifications that may be- 
come proposals for American Standards. This Committee should follow carefully the 
work of all other committees on engineering and may request any committee to investigate and 
prepare a report on the phase of motion picture engineering to which it is assigned. (File 70) 

F. E. Carlson, Chairman, General Electric Company, Nela Park, Cleveland 12, Ohio 

Chairmen of Engineering Committees 

Richard Blount L. W. Davee M. A. Hankins F. J. Pfeiff 

E. K. Carver H. H. Duerr H. J. Hood Leonard Satz 
M. H. Chamberlin E. C. Fritts D. E. Hyndman J. G. Stott 

C. G. Clarke R. L. Garman Rudolf Kingslake J. H. Waddell 

J. W. Cummings L. T. Goldsmith W. W. Lozier 

Members at Large 

Gordon Edwards E. W. Kellogg D. F. Lyman Otto Sandvik 

C. R. Keith G. T. Lorance 

Members Ex-Officio 

F. T. Bowditch V. O. Knudsen G. M. Nixon F. W. Sears 
L. A. Jones J. A. Maurer 

recommendations and prepare specifications on all phases of film equipment as used in the 
television broadcast stations. (File 75) 

F. N. Gillette, RTMA, Chairman, General Precision Laboratory, 63 Bedford Road, Pleasant- 
ville, N.Y. 

E. C, Fritts, SMPTE, Vice-Chairman, Eastman Kodak Co., 333 State St., Rochester 4, N.Y. 

A. J. Baracket L. C. Downes R. M. Morris C. L. Townsend 

Pierre Boucheron J. A. Maurer N. F. Oakley M. G. Townsley 

P. F. Brown H. C. Milholland R. C. Rheineck H. E. White 

Sydney Cramer G. C. Misener J. H. Roe 


FILMS FOR TELEVISION. To make recommendations and prepare specifications on all 
phases of the production, processing and use of film made for transmission over a television 
system excluding video transcriptions. (File 80) 

R. L. Garman, Chairman, General Precision Laboratories, Inc., 63 Bedford Road, Pleasant- 

ville, N.Y. 

M. R. Boyer H. R. Lipmaii R. M. Morris C. L. Townsend 

R. O. Drew G. C. Misener v R. C. Rheineck L. F. Transue 

Richard Hodgson Pierre Mertz H. J. Schlafly T. G. Veal 

R. Johnston H. C. Milholland N. L. Simmons H. E. White 

TELEVISION STUDIO LIGHTING. To make recommendations and prepare specifications 
on att phases of lighting employed in television studios. (File 85) 

Richard Blount, Chairman, General Electric Co., Nela Park, Cleveland 12, Ohio 

H. R. Bell H. M. Gurin Robert Morris W. F. Rockar 

A. H. Brolly Eric Herud R. S. O'Brien R. L. Zahour 

THEATER TELEVISION. To make recommendations and prepare specifications for the con- 
struction, installation, operation, maintenance, and servicing of equipment for projecting 
television pictures in the motion picture theater, as well as projection-room arrangements 
necessary for such equipment, and such picture-dimensional and screen-characteristic mat- 
ters as may be involved in high-quality theater-television presentations. (File 90) 

D. E. Hyndman, Chairman, Eastman Kodak Co., 343 State St., Rochester 4, N.Y. 
Ralph Austrian T. T. Goldsmith, Jr. Nathan Levinson L. L. Ryder 
G. L. Beers Nate Halpern W. W. Lozier Otto Sandvik 
F. E. Cahill, Jr. Richard Hodgson G. P. Mann Ed Schmidt 
James Frank, Jr. C. F. Horstman R. H. McCullough A. G. Smith 
R. L. Garman L. B. Isaac F. R. Norton E. I. Sponable 

E. P. Genock A. G. Jensen Harry Rubin J. E. Volkmann 
A. N. Goldsmith P. J. Larsen 

TEST FILM QUALITY. To develop and keep up to date all test film specifications, and to 
supervise, inspect and approve methods of production and quality control of all test films sold 
by the Society. (File 96) 

F. J. Pfeiff, Chairman, Altec Service Corp., 250 W. 57 St., New York 19, N.Y. 

R. M. Corbin Gordon Edwards Joseph Spray M. G. Townsley 

Russell Drew J. A. Maurer J. G. Stott 

THEATER ENGINEERING. To make recommendations and prepare specifications of engi- 
neering methods and equipment of motion picture theaters in relation to their contribution 
to the physical comfort and safety of patrons, so far as can be enhanced by correct theater de- 
sign, construction, and operation of equipment. (File 100) 

Leonard Satz, Chairman, Raytone Screen Co., 165 Clermont Ave., Brooklyn 5, N.Y. 
F. W. Alexa Charles Bachman James Frank, Jr. Ben Schlanger 

Henry Anderson E. J. Content Aaron Nadell Seymour Seider 

O. P. Beckwith C. M. Cutler E. H. Perkins Emil Wandelmaier 


Journal of the Society of 

Motion Picture and Television Engineers 


Color Television FRANK H. MC!NTOSH and ANDREW F. INGLIS 343 

Addendum: Recent Developments in Color Television 364 

Color Cathode-Ray Tube With Three Phosphor Bands 


A Magnetic Record-Reproduce Head M. RETTINGER 377 

Physical Principles, Design and Performance of the Ventarc High- 
Intensity Projection Lamps EDGAR GRETENER 391 

The High-Speed Photography of Underwater Explosions . . PAUL M. FYE 414 

A Heavy-Duty 16-Mm Sound Projector EDWIN C. FRITTS 425 

Interference Mirrors for Arc Projectors G. J. KOCH 439 

Engineering Committees Activities 443 


Questions and Answers in Television Engineering, by Carter V. Rabinoff and 

Magdalena E. Walbrecht Reviewed by Richard H. Dorf 444 

Reunions D'Opticiens, Tenues a Paris en Octobre 1946, Textes rassembl^s 
par Pierre Fleury, Andre Marshal et Mme. Claire Anglade, Institut 

d'Optique, Paris Reviewed by Dr. K. Pestrecov 445 

Photographic Instantanee et Cinematographic Ultra-Rapide, par P. Fayolle 

et P. Naslin Reviewed by John H. Waddell 445 


By Don Norwood 447 

New Members 448 

New Products 450 

Meetings of Other Societies 451 


Chairman Editor Chairman 

Board of Editors Papers Committee 

Subscription to nonmembers, $12.50 per annum; to members, $6.25 per annum, included in 
their annual membership dues; single copies, $1.50. Order from the Society's General Office. 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and 
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture and Tele- 
vision Engineers, Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General 
and Editorial Office, 342 Madison Ave., New York 17, N.Y. Entered as second-class matter 
January 15, 1930, at the Post Office at Easton, Pa., under the Act of March 3, 1879. 
Copyright, 1950, by the Society of Motion Picture and Television Engineers, Inc. Permission 
to republish material from the JOURNAL must be obtained in writing from the General Office of 
the Society. Copyright under International Copyright Convention and Pan-American Con- 
vention. The Society is not responsible for statements of authors or contributors. 

Society of 

Motion Picture and Television Engineers 

342 MADISON AVENUE, NEW YORK 17, N.Y., TEL. Mu 2-2185 

Earl I. Sponable 
460 W. 54 St. 
New York 19, N.Y. 

Peter Mole 

941 N. Sycamore Ave. 
Hollywood 38, Calif. 



Loren L. Ryder 
5451 Marathon St. 
Hollywood 38, Calif. 

Clyde R. Keith 
120 Broadway 
New York 5, N.Y. 

Robert M. Corbin 
343 State St. 
Rochester 4, N.Y. 

William C. Kunzmann 
Box 6087 
Cleveland 1, Ohio 

Fred T. Bowditch 
Box 6087 
Cleveland 1, Ohio 


Frank E. Cahill, Jr. 
321 W. 44 St. 
New York 18, N.Y. 

Ralph B. Austrian 
25 W. 54 St. 
New York 19, N.Y. 


Herbert Barnett 
Manville Lane 
Pleasantville, N.Y. 

Kenneth F. Morgan 
6601 Romaine St. 
Los Angeles 38, Calif. 

Norwood L. Simmons 
6706 Santa Monica Blvd. 
Hollywood 38, Calif. 


Charles R. Daily 
113 N. Laurel Ave. 
Los Angeles 36, Calif. 

John P. Livadary 
4034 Cromwell Ave. 
Los Angeles, Calif. 

William B. Lodge 
485 Madison Ave. 
New York 22, N.Y. 


Lorin D. Grignon 
1427 Warnall Ave. 
Los Angeles 24, Calif. 

Paul J. Larsen 
4313 Center St. 
Chevy Chase, Md. 

William H. Rivers 
342 Madison Ave. 
New York 17, N.Y. 


F. E. Carlson 
Nela Park 
Cleveland 12, Ohio 

George W. Colburn 
164 N. Wacker Dr. 
Chicago 6, 111. 

Edmund A. Bertram 
850 Tenth Ave. 
New York 19, N.Y. 

Malcolm G. Townsley 
7100 McCormick Rd. 
Chicago 45, 111. 

Edward S. Seeley 
161 Sixth Ave. 
New York 13, N.Y. 

R. T. Van Niman 
4501 Washington St. 
Chicago 24, 111. 

Color Television 


SUMMARY: The three major systems being considered by FCC are: (1) 
the CBS field sequential; (2) the RCA dot sequential; and (3) the CTI line 
sequential system. They are mutually exclusive in that different receiver 
circuit arrangements are required for each. Their principal characteristics 
will be described and related to the general problem of providing color tele- 
vision service to the public. The probable design factors upon which the 
FCC decision will be based, such as required bandwidth versus picture 
fidelity and resolution, and compatibility with existing black-and-white tele- 
vision systems will be covered in detail. 

THE ENGINEERING PROBLEMS which must be solved in developing 
a system of color television are exceedingly complex, and to 
discuss all of them in a single paper is obviously impossible. The 
most fundamental problem, however, which must be faced in the 
development of a broadcast color television system, is the trans- 
mission of a picture of maximum possible quality in a minimum 
bandwidth. All systems of color television now being proposed 
represent various approaches to this problem, and in order to under- 
stand these systems it is necessary to have a general understanding 
of its nature. The first part of this paper will, therefore, be devoted 
to the general subject of television systems as related to bandwidth. 


It is a well-known physical law that, with the type of amplitude 
modulation used for television broadcasting, the amount of informa- 
tion which can be transmitted is directly in proportion to the avail- 
able bandwidth. This law constitutes a most important limitation 
to television performance. The amount of information transmitted 
in an ordinary black-and-white television signal requires a video 
band of approximately 4 megacycles or about 400 times as great as 
that required for a reasonably high fidelity audio signal. This video 
bandwidth is translated by the modulation process to radio-fre- 
quency bandwidth, and with single sideband amplitude modulation 
6 megacycles of space in the radio-frequency spectrum is required to 
transmit a 4-megacycle video signal plus the accompanying audio 
channel on a separate transmitter. 

PRESENTED: April 24, 1950, at the SMPTE Convention in Chicago. 


344 MclNTOSH AND iNGLis October 

Because of the serious shortage of available spectrum space, the 
problem of utilizing it to its utmost is probably the most serious 
facing the television engineer. A tremendous amount of ingenious 
research has been devoted to this problem, some of the most im- 
portant results of which will be discussed here. 

In a black-and-white television picture the information to be trans- 
mitted is the variation in brightness of various portions of the picture. 
The number of independent brightness values per second which can 
be transmitted is equal to twice the highest frequency passed by the 
system. Each one-half cycle period of time is sometimes called a 
Nyquist interval. With the standard monochrome television 
broadcast system, the highest frequency is 4 megacycles and the 
Nyquist interval is one-eighth of a microsecond. This means that 
the maximum number of independent brightness values (which is 
usually expressed as the number of picture elements) will be 8,000,000 
per second. This number represents the product of the number of 
picture elements in a single picture and the number of pictures which 
are transmitted each second. It is quite apparent, therefore, that 
the smaller the number of pictures per second, the larger will be the 
number of picture elements and correspondingly the greater the 
resolution of the picture. For this reason it is desirable to transmit 
as few pictures per second as possible. The limit to the minimum 
number of pictures which can be transmitted per second is set by the 
factors of flicker and the blurring of moving objects. Of these, the 
problem of flicker is the more important and has received the greater 
amount of attention. 

With standard monochrome television, the problem of flicker has 
been reduced by using an interlaced scanning system. In this 
system alternate lines are scanned during a period of one-sixtieth of 
a second. During the next one-sixtieth of a second, the other set of 
alternate lines is scanned so that the complete picture is transmitted 
in a period of one-thirtieth of a second. Because of the method of 
interlaced scanning, however, it has been determined that the large- 
area flicker is probably no more objectionable than would be the case 
if the complete picture were transmitted 60 times a second, though 
a less visible "interline flicker" is introduced, particularly when the 
received picture is bright. This 2-to-l system of interlace, therefore, 
doubles the number of picture elements which can be transmitted 
within a given bandwidth. 

In a 4-megacycle bandwidth capable of transmitting 8,000,000 
picture elements per second and picture repetition rate of 30 per 
second, the maximum theoretical number of picture elements per 


picture would be 266,000. This theoretical maximum is not reached 
in practice, however. There is a reduction of approximately 14% 
due to the time consumed during the horizontal blanking interval at 
the end of each line, and a reduction of approximately 8% due to the 
vertical blanking interval at the end of each field. From these con- 
siderations alone the picture elements which theoretically can be 
transmitted in a black-and-white picture are approximately 79% 
of 266,000 or about 210,000. There is an additional reduction factor 
which is due to the fact that vertical variations in brightness cannot 
occur at arbitrary positions but can occur only from one line to the 
next. Since the position of the scanning lines may not coincide ex- 
actly with the variations in brightness of the picture, the effective 
vertical resolution is not equal to the number of visible scanning lines. 
A factor of approximately 0.7 is normally used to express the reduc- 
tion in resolution from this source. When these factors are all com- 
bined, the total number of picture elements which can be practically 
resolved is found to be about 147,000. This figure is equal to the 
product of the number of lines which can be resolved in the vertical 
direction which is about 338 (525 X 0.92 X 0.7) and the horizontal 
direction which is about 437 (63.5 /*sec (microseconds) X 0.86 X 8 
lines//isec). When allowance is made for the 4-to-3 aspect ratio, 
the number-of-lines resolution in a horizontal distance equal to the 
height of the picture is computed to be 328 which is almost identical 
with the vertical. 


Since vertical interlace can be used to double the resolution of a 
picture for a given bandwidth, it is suggested that the resolution 
might be doubled again by using horizontal interlace also. This 
technique has been discussed at length in recent testimony before 
the FCC (Federal Communications Commission) in connection with 
the proceedings on color television and also in the literature.* 

The fundamental objective to be achieved by this type of hori- 
zontal interlace is to cut the rate of picture transmission in half, 
thereby doubling the number of picture elements which can be trans- 
mitted in each picture. To accomplish this, the waveform to be 
transmitted is first sampled at approximately twice the highest fre- 
quency which can be transmitted. This is illustrated in Fig. 1 
which shows a 6-megacycle sine wave sampled at the rate of 8 mega- 

* Wilson Boothroyd, "Dot systems of color television," Electronics, Pt. I, vol. 
22, no. 12, pp. 88-92, Dec. 1949; Pt. II, vol. 23, no. 1, pp. 96-99, Jan. 1950. 




cycles, that is, a train of pulses having a repetition rate of 8 mega- 
cycles is amplitude modulated so that the height of each pulse is 
equal to the amplitude of the signal at that instant. In the television 
application, this sine wave would represent the output of the camera 
along one line of the picture. During the succeeding frame, assum- 
ing that no motion has occurred in the picture, this same sine wave 
will occur, and it is again sampled at the rate of 8 megacycles, but 
with the sampling points shifted by one-half cycle as indicated by the 
dotted lines. When pulses, amplitude modulated by a signal of 
greater frequency than 4 megacycles, are passed through amplifier 






Figure 1. 

circuits having a bandwidth of only 4 megacycles, mathematical 
analysis shows that the only frequencies passed will be the d-c com- 
ponent and a component equal in frequency to the difference between 
the signal and the sampling frequencies. Thus in this case, the trans- 
mitted signal for both frames will consist of a d-c component plus a 2- 
megacycle frequency. These 2-megacycle components, however, will 
be 180 deg out of phase as indicated in Fig. l(b). 

At the receiver, these 2-megacycle components are again sampled 
in synchronism with the transmitter, thus reconstructing the trans- 
mitter samples as indicated in Fig. 1 (b) . At this point, a number of 
mechanisms could be employed to present this information on the 


kinescope perhaps the most obvious of which would be to use these 
narrow samples to control the magnitude of a wider pulse occupying 
approximately ^{Q /zsec in the signal applied to the kinescope. This 
would result in a series of dots being produced on the screen with the 
spaces between the dots being rilled in alternate fields. In this way, 
it would be possible to have independent signal components occurring 
at the equivalent rate of 16 million per second which is equivalent to 
a bandwidth of 8 megacycles without horizontal interlacing. A 
more careful analysis indicates, however, that this process is not 
necessary. Instead, it is only required to pass the sample pulses 
through a filter, passing all frequencies up to but not including the 
sampling frequency. When this is done, it can be shown that two 
frequency components will appear in the output: first, the original 
signal frequency, 6 megacycles in this case; and second, the differ- 
ence frequency, 2 megacycles in this case. In the succeeding frame, 
the difference frequency will again appear but 180 deg out of phase 
with respect to the first frame. If the sample output from the first 
frame is stored and added to the output of the second frame, the 
difference frequency components will cancel out, thus leaving only 
the signal frequency component. A cathode-ray tube with a phos- 
phor of sufficient persistence when combined with the natural per- 
sistence of the eye provides a method for storing and adding these 
signals, the only limitation being that the negative values of bright- 
ness cannot be physically achieved. Consequently, if the output of 
the sampling circuit is applied directly to the control grid of the 
kinescope through a filter as described above, the signal components 
will be automatically produced correctly. As in the case of line 
interlace a new form of flicker will appear, "inter-dot" flicker, which 
is most visible with vertical arrays of dots, or vertical or diagonal 
lines, in the picture. 

In the example given above, a specific frequency component of 
6 megacycles was used for illustration. The following general 
description can be made for other frequencies. Assuming again that 
the transmission system will pass components from to 4 megacycles 
and that the sampling rate is 8 megacycles, signal components from 
to 4 megacycles will be unaffected by the sampling process and will 
be transmitted in the usual manner. Signal components from 4 to 8 
megacycles will beat with the 8 megacycle sampling frequency and 
will appear in the transmitted signal as a difference frequency. 
These difference frequencies will likewise follow in the range from to 
4 megacycles. However, when resampled at the receiver, they are 
distinguishable from the original signal components by the fact that 

348 MclNTosH AND iNGLis October 

on alternate frames, they will be 180 deg out of phase so that when 
they are sampled at the receiver the original signal frequency is 

When transmitting a color picture, in addition to variations in 
brightness, variations in color must also be transmitted. It might be 
thought, therefore, that the bandwidth required to transmit three 
primary colors would be three times as great as that required for 
black-and-white. Early experiments in color television followed this 
theory. At that time two methods were under consideration: the 
field-sequential system and the simultaneous system. In the simul- 
taneous system, the three color signals were sent over separate chan- 
nels so that the total bandwidth required was nearly three times that 
required for single black-and-white transmission. (The total was 
not quite three times as great, inasmuch as a reduced bandwidth was 
used for the blue signal; this could be done because the eye is not so 
sensitive to detail in blue.) In the field-sequential system, the suc- 
cessive fields were transmitted cyclically in the three primary colors, 
but when this was done it was found necessary to increase the field 
rate to approximately three times the black-and-white rate in order 
to avoid objectionable color flicker. This resulted in a nearly three- 
fold increase in bandwidth to achieve the same picture resolution. 

A closer analysis of the situation, however, indicated that this in- 
crease in bandwidth for the transmission of color is not inherently 
required. If a scheme could be worked out whereby a smaller num- 
ber of pictures could be transmitted per second without objectionable 
flicker, the resolution of standard black-and-white could be main- 
tained. Two of the three systems of color television now proposed 
to the FCC attempt to slow down the rate of picture transmission 
without a corresponding increase in flicker. These methods will be 
described later in connection with the individual systems proposed. 


The simultaneous method of color television has now been aban- 
doned for broadcast television because of its excessively large band- 
width requirements, and all three systems now proposed are sequen- 
tial in nature; that is, the color information is sent in cyclical se- 
quence and the retentivity of the human eye is employed to pro- 
duce the illusion of a continuous color picture. The essential dif- 
ference between the three systems lies in the color switching rate. 
This is illustrated in Fig. 2. On the left side of this figure is shown 
in schematic form the field-sequential system proposed by CBS 
(Columbia Broadcasting System). In this system, a complete field is 




transmitted in one color. The following field is sent in the second 
color and the following field in the third, and so forth. 

In the center of the figure is illustrated the line-sequential system 
proposed by CTI (Color Television, Inc.). In this system the color 
is changed at the end of each line so that each field contains a third 
of its scanning lines in each primary color. 










Figure 2. 

On the right side of the figure is shown the dot-sequential system 
proposed by RCA (Radio Corporation of America). In this system, 
the color is switched many times during each line. The rate currently 
proposed by RCA is to complete the color cycle in a period correspond- 
ing to a frequency of 3.6 megacycles. 

350 MclNTOSH AND iNGLis October 

The extreme difference in the color switching rate between the 
RCA system on one hand and the CBS on the other is such that the 
terminal apparatus suitable for the three systems differs consider- 
ably. These apparatus differences have tended to obscure the more 
fundamental difference in the color switching rate, which is the basic 
distinction between the three systems. 


As stated above, CBS employs the field-sequential system wherein 
a complete field is transmitted in one color and the color change is 
made during the vertical blanking period. In order to avoid flicker, 
it has been found necessary to increase the field rate from the standard 
60/sec employed by black-and-white television to 144/sec. Two 
fields are required to transmit the entire picture in a single primary 
color and 6 fields are required to transmit a complete color picture. 
There are, therefore, 24 color pictures transmitted each second, 
corresponding nicely to the current 35-mm motion picture frame rate 
(although it must be admitted that such "compatibility" is of far less 
importance than "compatibility" with present black-and-white 
television, discussed below) . 

Because the field rate has been increased from 60 to 144/sec 
with no change in the interlace pattern, there is an inevitable decrease 
in resolution in the ratio of 60 to 144. This decrease in resolution is 
divided between the vertical and horizontal directions. The vertical 
resolution is reduced by decreasing the number of scanning lines 
from 525 to 405, a reduction of approximately 27%. A single line 
occupies 34.4 jusec as compared to 63.5 for present black-and-white. 
The number of Nyquist intervals in this period is, of course, reduced 
in the same ratio so that the horizontal resolution is reduced by 
approximately 46%. 

In order to increase the horizontal resolution of the picture, a 
system of horizontal interlace somewhat similar to that described 
above, has been proposed by Dr. Peter Goldmark, inventor of the 
CBS system. This will restore the horizontal resolution to approxi- 
mately the same value which exists for standard monochrome. It 
should be pointed out that this signal can be received without sub- 
stantial degradation by a receiver not equipped with a sampling cir- 
cuit. The resolution obtained in this case, of course, is not im- 
proved by the interlace process. 

Since the CBS field rate is different from that employed by standard 
black-and-white, it is not a "compatible" system; that is, standard 
black-and-white receivers are unable to receive the CBS transmissions 


even as black-and-white pictures without the addition of a special 
scanning adapter which will cause the receiver sweep circuit to operate 
in synchronism with the transmitter at the increased rate necessary. 
Messrs. Chapin and Roberts of the FCC Laboratory Division have 
developed an automatic adapter for the CBS system which is said 
to make it possible for the receiver sweep circuits to operate auto- 
matically in synchronism with the transmitter no matter which field 
rate is being transmitted. The installation of this device in a new 
receiver would, of course, increase its cost somewhat, and to adapt 
sets which have already been installed would cost somewhat more. 
This fact, plus the operational confusion which would result with dual 
standards, has perhaps been the source of the most severe criticism 
of the CBS system. 

The CBS system employs the slowest color switching rate of the 
three and because of this it is the only system which can accomplish 
it by mechanical means. It has somewhat unfairly been accused of 
being inherently mechanical. This accusation is not fair inasmuch as 
any of the electronic terminal equipment, as employed by the other 
systems, could be used by CBS when suitably modified. Moreover, 
the possibility of color switching mechanically leads to certain 
definite advantages over all-electronic methods. 

Some of the camera pickup equipment which could be used by CBS 
is illustrated in Fig. 3. At the top of this figure is shown the mechani- 
cal method which has been used almost exclusively by CBS in its 
experiments. With this device, a disc divided into segments covered 
with various color filters is rotated in front of the pickup tube. The 
shape of the filter segments and the speed of rotation of the disc is 
chosen in such a way that during each field a filter of given color will 
be in front of the pickup tube. 

In the second drawing shown as Fig. 3(b), a system somewhat 
similar to that employed by CTI, has been adapted for use with the 
field-sequential system. Three lenses placed one above the other 
and each provided with a filter for one of the three primary colors 
focus three images on the light sensitive surface of the pickup tube. 
The horizontal sweep on the pickup tube is made to operate at its 
normal rate, but the vertical sweep operates at one-third the field 
frequency so that during one vertical period all three images are 
scanned. If th6 vertical sweep is perfectly linear and the images are 
properly positioned, a suitable field-sequential signal is created. 

Figure 3(d) shows the camera employed by RCA. Here a single 
lens is used and the light emerging from the lens is made to pass 
through crossed dichroic mirrors. These mirrors transmit green 




light directly to the center tube, one of them reflects red light to one 
of the tubes on the side while the other reflects blue light. Thus each 
of the three pickup tubes is illuminated constantly with light from 
one of the three primary colors. The outputs of the sweep circuits 
are applied to the deflection coils on the three tubes connected in 
parallel so that identical voltages are applied to each. Thus three 
signals are available constantly, one from each of the three tubes. 
In order to adapt this camera to the field-sequential system, it is only 
necessary to employ a gating circuit which will pass the output of 





Figure 3. 

each of the three camera tubes in sequence, switching at the end of 
each field. 

In its demonstration CBS has used a mechanical system almost 
exclusively. This is probably due to the fact that only one tube is 
required and the fact that optical and electronic registration problems 
are avoided. The problems of registration are undoubtedly the 
most serious for electronic methods of color television and this is 
neatly solved by a mechanical disc. 

The possible types of receivers which might be used with the 
CBS system are shown in Fig. 4. Figure 4 (a) shows the mechanical 
method which is analogous to the system employed at the camera. 




Here a disc divided into filter segments is rotated in front of the 
kinescope. This disc is synchronized in such a way that the red 
filter is in front of the picture tube when the red signal is being trans- 
mitted, the blue for the blue signal, and so forth. The eye then 
blends all three impressions together as a single colored picture. 









Figure 4. 

Figure 4(b) shows a scheme similar to that shown on Fig. 3(b) for 
the camera. The phosphors on the face of the tube are deposited in 
three strips one above the other, each phosphor being chosen in such a 
way that the light emitted from it corresponds approximately to one 
of the primary colors. (In practice fairly good green and blue 
primaries can be achieved. The red primary at present contains too 
much yellow-green and a negative yellow-green filter is normally 
used.) The three images thus produced on the face of the tube are 

354 MclNTosH AND INGLIS October 

superimposed by means of a lens system upon a projection screen, 
either reflecting or translucent, thus producing a color picture. 
Figure 4(d) shows a color receiver developed by RCA but which can 
also be adapted for use by CBS. Here three kinescopes are used. 
One of the tubes is viewed directly through a dichroic mirror while 
the virtual images of the other two are superimposed upon this direct 
image by means of the dichroic mirrors. These mirrors also provide 
the necessary filtering action so that one of the tubes provides an 
image in each of the three primary colors. The two virtual images 
superimposed on the face of the three tubes provide the illusion of a 
complete color picture. To adapt this receiver for the CBS system, 
it is only necessary to gate the input to the tubes so that the red tube 
operates with the red signal and the green with the green signal, and 
so forth. 

Figure 4(e) shows the direct-view tube recently announced by 
RCA. The operation of this tube will be described later. This 
tube although developed by RCA for use in connection with its system 
could be adapted with ease to the CBS system. 

Of these methods, CBS has used extensively only the mechanical 
system. The direct-view color tube has just been announced and 
has not been available for experimentation by companies other than 
RCA. As compared with the other two electronic methods, the 
mechanical system has the advantage of providing a means of pro- 
ducing a direct-view color picture at relatively low cost and with no 
registration problem. As in the case of camera equipment, the 
problem of registration is a very difficult one with electronic systems. 
On the other hand, the color disc has certain disadvantages, among 
which are its awkward size, and the fact that its use is limited to 
kinescopes of approximately 10-in. diameter due to the peripheral 
speed and awkward size of the whirling disc. 

When the receiver and transmitter are operating on the same power 
line frequency, a synchronous motor can be used to operate the color 
disc. This will automatically stay in synchronism with the trans- 
mitter once it is set. In the case of the network programs where the 
problem of nonsynchronous power supply arises, it is necessary to 
transmit a phasing signal to keep the two in step. This signal as 
proposed by CBS consists of a series of impulses transmitted during 
the vertical blanking interval. By suitable circuitry, this impulse 
is made to operate the motor at the receiver in synchronism with the 
transmitter. The exact method to be used by CBS has not been 



The line-sequential system proposed by CTI switches colors be- 
tween (successive) lines during the horizontal blanking interval and 
in a given field, the lines are cyclically scanned in the three primary 
colors. Three possible sequences can be followed from frame to 
frame: a given line can always be traced in the same color, it can be 
traced in two of the primary colors or it can be traced in all three of 
the primary colors. Since 525 is (integrally) divisible by three, if a 
cyclic order were continuously maintained in color switching, the 
first situation would automatically follow; for example, the first line 
would always be in red, the second in green and the third in blue, and 
so forth. If during the vertical blanking interval, the cycle is upset 
so that, for example, one of the green lines is omitted, the "phase" 
of the color cycle will be shifted by one-third of a cycle. When this 
is done, a line which is scanned in red the first time can be scanned 
in blue the second and in green the third. This color shift is accom- 
plished during the vertical blanking interval so that the shift in the 
color cycle is not noticeable. If one shift is performed every three 
fields, each line will eventually be scanned in two of the primary 
colors. This method of operation has, therefore, been termed by 
CTI, the "single-shift." If such a color shift is made during two out 
of three vertical blanking periods, each line will be scanned in all three 
primary colors and this method of operation has been termed by CTI 
as the "double-shift." 

In a double-shift system, each line is scanned in all three colors 
and the color picture is, therefore, complete. With a single shift, 
one color is missing from each of the three lines. This has two effects : 
The vertical resolution of the picture is reduced by some amount, the 
exact reduction depending somewhat on the degree of saturation of 
the picture. If the picture consists of a single primary color, one- 
third of the lines will be black and the vertical resolution will be re- 
duced by one-third. The reduction of vertical resolution for colors 
consisting of a mixture of primaries will be somewhat less. The 
second effect is the fact that the black lines will tend to show in the 
simple primary colors, and that the lines will be colored even when a 
white object appears in the picture. This is probably not a serious 
defect inasmuch as the eye is not sensitive to colors in such small 
areas as a single scanning line and the integrated effect of all the lines 
produces the proper color sensation at the eye, at least at a great 
enough viewing distance. With no shift in the color cycle, the same 
effects will occur except to a greater degree. The reduction in verti- 




cal resolution with no color shift is so great that it has not been 
seriously proposed. 

The choice of a proper scanning sequence is very important for the 
operation of the line-sequential system and a large number of com- 
binations have been tried. Two of these are shown in Fig. 5, the 
first of which is a single-shift pattern in which each line is scanned in 
two of the primary colors while the second is a double-shift pattern 
in which each line is scanned in all three of the primary colors. 

A third scanning sequence has recently been demonstrated by 
CTI which is described as the "interlaced shift." This is a double- 


























































































Figure 5. 

shift pattern in that each line is ultimately scanned in all three colors 
with six fields being required for a complete color picture. It 
departs from the standard interlace in that odd lines are scanned 
during three successive fields and the even lines on the next three. 
The purpose of this shift is to mitigate line crawl. This is accom- 
plished by employing a sequence wherein continuous upward and 
downward motions of lines of each color occur simultaneously. 
While either of these motions can be seen if specifically looked for, 
the effects of the two cancel out as soon as the observer concentrates 
on the subject matter rather than the line structure. This sequence 
has been described by CTI as the best of all it has tried. 


The line-sequential system as advocated by CTI employes the 
same number of fields per second, the same number of lines, and the 
same 2-to-l interlace employed by standard black-and-white tele- 
vision. This, therefore, is a compatible system in that ordinary 
black-and-white receivers can receive the transmission in black-and- 
white without any changes or adjustments necessary. 

Since the period occupied by one line is the same as for standard 
black-and-white transmission, the number of Nyquist intervals in 
one line is the same and the horizontal resolution is likewise identical. 
If the double-shift pattern is used, the vertical resolution is likewise 
identical while if the single shift is used, the vertical resolution is 
reduced by some factor which may be as high as one-third. 

Any of the electronic cameras described in connection with the 
CBS system can be used with the CTI. The camera shown in Fig. 
3(b) is modified by placing the three images side by side on the 
photosensitive surface rather than one above the other. The vertical 
sweep frequency is at the normal rate but the horizontal sweep is 
reduced to one-third of the standard value. Thus as the electron 
beam scans across the mosaic it strikes successively, a red, green and 
blue image area thus providing modulation for a red line, a green line 
and a blue line in order. At the end of the field any desired change in 
shift in the color cycle can be made so that the colors in the next field 
will be scanned in the desired order. 

The three-tube camera can also be used (Fig. 3(d)). Here it is only 
necessary to switch the output of the three cameras in such a way 
that the red signal is transmitted for one line, the green signal for the 
next, the blue for the next, and so forth. 

So far in its demonstrations, CTI has employed the single-tube 
type of pickup. This, of course, is considerably cheaper than the 
three-tube system employed by RCA. However, the problem of 
making the sweep sufficiently linear so that the three images can be 
maintained in proper registry has been quite bothersome, and it 
may be that the three-tube camera can be operated with greater ease. 
Also, the resolving power of existing orthicons provides a limitation 
to the resolution of the picture which can be obtained when three 
small images are placed side by side on the photosensitive surface. 

Any of the electronic receivers described in connection with the 
CBS system can be used for the CTI line-sequential system. In the 
case of the single-tube projection receiver, the three colored images 
are placed side by side rather than one on top of the other (Fig. 4(c)). 
The vertical swqep operates at the normal rate while the horizontal 
operates at one-third of this value. With the RCA three-tube 

358 MclNTosH AND iNGLis October 

receiver (Fig. 4(d)) it is only necessary to gate the input to the three 
tubes so that the red signal actuates the red tube for one line, the 
green for one line and the blue for another, and so forth. The RCA 
direct-view color tube (Fig. 4(e)) can also be used with the CT1 line- 
sequential system. None of these tubes have been available to CTI 
as yet for experimentation. However, it seems apparent that the 
simplicity and direct-view characteristics of this will make it or some 
direct-view tube the ultimate answer to the receiver problem for the 
line-sequential system. 

When the double shift is used, six fields occupying one-tenth of a 
second are required to trace every line in each of the primary colors. 
It is by thus slowing down the rate of transmitting the complete color 
information that CTI is able to maintain the resolution of black-and- 
white and transmit color in addition. With the field-sequential 
method, slowing down the picture repetition rate to this speed has 
resulted in very objectionable flicker. By switching colors at the 
end of each line, the flickering elements are broken up into lines rather 
than fields and are thus very much less noticeable to the eye. The 
crucial question in connection with the line-sequential system is 
whether this rate of color switching is sufficiently rapid or involves a 
small enough area to reduce flicker to an unobjectionable value. 
With certain scanning sequences and with certain types of subject 
material an objectionable line crawl or interline flicker seems to occur. 
Critics of this system contend that this is an inherent flaw in the 
system which cannot be solved by any apparatus refinement. 
Whether or not this is true has not been definitely established and, 
as indicated above, it is probably upon this factor that the success or 
failure of the line-sequential system will depend. 

When the single shift is employed, a cycle is completed in four 
fields or one-fifteenth of a second. As might be expected, such a 
pattern is less susceptible to line crawl or flicker than the double 
shift. Again, no definitive answer has been determined with regard 
to the seriousness of the flicker problem with the single shift. 

As compared with the RCA system, to be described later, the CTI 
system has the obvious advantage of much less complexity, plus the 
fact that it can operate in a system of reduced bandwidth such as 
existing coaxial cable with the natural loss in resolution but with no 
loss in color values. The main question with regard to this system is 
whether this simplicity has been achieved at too great a loss in per- 
formance. Until the question of interline crawl or flicker and its 
effect on picture quality has been definitely determined, it will be 
impossible to say whether or not this is the case. 



The RCA color television system employs the most rapid color 
switching rate of the three sequential systems which have been pro- 
posed. For this reason, as might be expected, it is the most complex 
of the three, both with respect to the studio and receiving equip- 
ment. This system is, indeed, a marvel of ingenuity. 

In its present form, the RCA color camera consists of three pickup 
tubes arranged as shown in Fig. 3(d) which provide simultaneously 
red, green and blue electrical signals. Each of these signals is first 
passed through a filter which separates the components below and 
above 2 megacycles. The high-frequency components from 2 to 4 
megacycles are combined and are transmitted without color separa- 
tion. This procedure is justified by RCA on the grounds that the 
eye is not sensitive to colors in the fine detail represented by the high- 
frequency components in the picture; consequently, it is not neces- 
sary to transmit this detail in color. This has been termed the 
"mixed highs" principle by RCA. 

The three color signals containing frequency components up to 2 
megacycles are then sampled at the rate of 3.6 megacycles, that is, 
during the period corresponding to one cycle of this frequency all 
three colors are sampled. The amplitude of each pulse sample is 
proportional to the corresponding magnitude of each of the primary 
colors in this portion of the picture. The sample pulses are then 
combined and passed through a filter which removes all frequency 
components above 4 megacycles. As a result, the 3.6 megacycle 
sine wave is produced as shown in Fig. 6 (a). Examination of this 
figure indicates qualitatively that the phase of the resulting wave 
will be determined by the predominant color in the picture, the 
magnitude of the wave will depend on the difference between the high- 
est and lowest samples, while the d-c level of the wave will depend on 
the average level of the three components. More precisely, it may be 
said that the hue of the picture is determined by the phase of the 
color wave, the saturation of the color is determined by the ampli- 
tude of the color wave, and the brightness of the picture is determined 
by the d-c content of the wave. 

The color wave is combined with the "mixed highs" signal and 
both are then fed to the transmitter input and the signal is trans- 
mitted in the usual fashion. 

At the receiver, the converse of this action occurs. The trans- 
mitted wave is first amplified and detected in the usual manner. 
The resulting video signal is then sampled in synchronism with the 




transmitter and as a consequence, three sets of sampling pulses are 
made available one for each color. These pulses are then passed 
through a low pass filter which results in a 3.6-megacycle sine wave, 
the amplitude of which is proportional to the amplitude of the 
sample pulses. This process is illustrated in Fig. 6(b). These sig- 
nals are then combined with the "mixed highs" and the resulting 
signals are ready for application to the kinescopes or color tube. 






/IN /IK /t\ /I\ /l\ 


/IN /IN /IK /T\ /IN 


Figure 6. 

In the case of the three-kinescope receiver illustrated in Fig. 4(d), 
each of these signals is applied to the corresponding kinescope grid. 
The sinusoidal form of this signal results in a series of dots being 
produced on the kinescope and when these three sets of dots are 
combined, the result is the dot-sequential color presentation illustrated 
in Fig. 2. 

In the case pf the direct-view color tube recently demonstrated by 




RCA, the three signals are all applied to a single tube in a manner 
which will become apparent later when this tube is described. 

In order that each portion of the screen be excited eventually with 
all of the three colors, it is necessary to shift the phase of the color 
cycle so that on succeeding frames, each portion of the screen will be 
excited sequentially in the three colors. If there were no overlap 
between the dots, it would require three frames in which to do this; 
however, because the width of the dots is about one and one-half 
times the distance between them, it is possible to cover the entire 


V* G R E 


v B G 


^ B G 


I G R ' 


^ G R [ 


B G 













B G 












B *> 






















B G 











B G 












B <Ss 









Figure 7. 

area with three colors with two frames so that 15 complete color pic- 
tures are sent in each second. The order in which this is accom- 
plished is shown on Fig. 7. 

It is quite apparent that the transmission of color by the RCA 
system depends on accurate phasing of the sampler in the receiver 
with respect to the color wave in the transmitted signal. Any phase 
shift in the sampling oscillator with respect to this color wave will 
result in false colors being reproduced at the receiver. In its initial 
demonstrations before the FCC, RCA employed the trailing edge of 

362 MclNTOSH AND iNGLis October 

the horizontal synchronizing pulse to phase the sampling oscillator 
in the receiver. This did not appear to provide a sufficiently precise 
method of timing this circuit inasmuch as the colors during those 
demonstrations shifted constantly. Since that time, RCA has in- 
corporated a short train of 3.6-megacycle oscillations on the "back 
porch" of the horizontal blanking pedestal. m The receiver sampling- 
oscillator is made to synchronize in frequency and phase with this 
burst of 3.6 megacycles of energy. With this arrangement any shift 
in the phase of the color wave due to the circuits through which it 
must pass will likewise shift the phase of this "burst" of energy and, 
accordingly, there is no change in the relative phase of the sampling 
oscillator and the color wave. This device seems to have been quite 
successful inasmuch as the color stability shown in pictures in later 
demonstrations was quite good. 

As compared with the other two systems, the RCA system is said 
to be completely compatible with monochrome standards, to have 
geometric resolution equal to monochrome, and by breaking each 
line into dots, to avoid the interline flicker problem which is claimed 
by RCA to be the inherent flaw in the line-sequential method. On 
the other hand, this system has been criticized for its registration 
problem (which is present in any all-electronic system), its relative 
complexit}^ and the fact that the signal cannot be transmitted over a 
2.7-megacycle coaxial cable. RCA claims to have rem'oved this 
last objection by developing a system for sampling the picture at a 
2.4-megacycle rate for transmission over the coaxial cable and re- 
sampling the output of the cable at the 3.6-megacycle rate for trans- 
mission over the air. The RCA system is also claimed by its critics 
to be more susceptible to noise and interference from other stations 
than is the case with the other types of color transmission. 


It has been quite generally agreed by most engineers who have 
testified concerning color television systems that, regardless of which 
system is adopted, the ultimate method of presentation will be a 
direct-view color tube. It is quite natural then, that the recent 
demonstration of the RCA color tube was the object of extreme 
interest on the part of all industry engineers. It was quite generally 
conceded by all who saw the demonstration that this tube represented 
a remarkable engineering feat. While the tube, admittedly, has con- 
siderable room for improvement, it can definitely be said that a direct- 
view color tube is in sight for commercial use. To date, the details, 


particularly of the manufacture, of this tube have not been divulged 
by RCA, but a general description can be given. 

Of the two types demonstrated, one contains three electron guns 
and the other only one. In both types, the color screen consists of 
an array of small, closely spaced phosphor dots arranged in tri- 
angular groups, each group containing a green emitting dot, a red 
emitting dot and a yellow emitting dot. In the tube demonstrated, 
there were 351,000 dots in all with 117,000 dots for each color. Inas- 
much as a good red phosphor has not yet been developed, it is neces- 
sary to place a minus yellow-green filter in front of the screen in 
order to restore proper color balance to the picture. 

About one-half inch behind the color screen is placed a metallic 
mask which contains 117,000 holes or one hole for each of the tri- 
angular dot groups. This hole is so positioned with respect to its 
associated dot group that the difference in the angle of oncoming 
beam determines the dot which will be excited and consequently which 
will be seen by the viewer. 

In the case of the three-gun tube, the three guns are placed in a 
bundle as shown in Fig. 4(e) in the neck of the tube with their axes 
converging at the plane of the metallic screen. Because of the 
difference in angle at which beams from these three guns will strike 
the holes, each gun will excite the dots of one color only. Thus, it is 
necessary only to apply the red signal to one of the guns, the blue 
signal to another and the green signal to the third in order to provide 
a color picture. It is quite apparent that this tube could be used not 
only with the RCA dot-sequential but also with the field-sequential 
and line-sequential systems. 

In the case of the single-tube kinescope, the beam is magnetically 
displaced, and this displacement is rotated so that, in effect, it oc- 
cupies in time sequence, the positions of the three guns in the three- 
gun kinescope. As the beam displacement rotates, the angle at 
which the beam enters the holes in the perforated screen likewise 
changes and the different dots are excited in sequence. This tube 
can likewise be adapted to the dot-, line- or field-sequential systems 
by rotating the beam displacement at the dot, line or field rates. 
While the beam is being rotated in this manner, its intensity is modu- 
lated in accordance with the amount of color information present in the 
signal for the particular primary color being transmitted at the 
moment, this action automatically carrying out the sampling function. 

It is apparent that with this tube, there will be 117,000 picture 
elements in the reproduced picture. This is a reduction in resohi- 

364 MclNTOSH AND iNGLis October 

tion as compared to what can be obtained in a 4-megacycle band- 
width and one of the lines of development now being pursued by RCA 
is the construction of a tube with a larger number of picture elements. 


The attempt here has been to distinguish between the three systems 
on the basis of the color switching rate rather than on the terminal 
equipment which may have been employed by the three proponents 
in their demonstrations. This has been done because it is a problem 
of choosing from among the three color switching rates that is before 
the Federal Communications Commission in setting engineering 
standards, rather than the problem of determining which company 
has developed the best apparatus. Once this decision is made by the 
FCC, it can be expected that the entire industry will turn to the 
problem of developing and improving transmitting and receiving 
equipment for the particular system which has been chosen . Tremen- 
dous developments have occurred within the last nine months in the 
field of color television, and if these continue at their present rate, 
it can confidently be expected that it will be a commercial reality 
within a very few years. 

ADDENDUM, October 7, 1950 
Recent Developments in Color Television 

Since the record of the color television proceedings was closed by the FCC, 
two very interesting alternative systems have been publicly described. 
These are the frequency interlace system developed by the General Electric 
Co. and the "Uniplex" system developed by CTI. Neither of these has been 
tested to date to the extent of constructing complete transmitting and re- 
ceiving apparatus. However, considerable study has been made of the 
theoretical aspects of each, and certain critical points have been tested under 
simulated conditions. 

The GE frequency interlace system makes use of the fact that most of the 
energy in the video waveform is presumed to occur at harmonics of the line 
frequency of 15,750 cycles/sec. If each line were identical, the energy would 
be concentrated at these discreet frequencies. Due to the fact that the 
waveform changes from line to line, the frequency spectrum is spread out 
somewhat, and as a consequence it is believed by GE that picture information 
is concentrated in regions of the spectrum lying near the harmonics of the line 
scanning frequency and occupying about 54% of the total spectrum. Since 
the other 46% of the spectrum space is not occupied, it is theoretically avail- 
able for transmission of additional information. 

While research has not yet been carried to the extent of determining the 
best way in which to utilize this unused spectrum space for transmission of 


color information, GE has suggested that it might be accomplished in the 
manner described below. 

At the camera the three color signals are produced simultaneously. This 
could be accomplished, for example, with the RCA color camera. Each 
signal may contain frequencies extending to 4 megacycles. However, since 
the higher frequency components in the red and blue signals do not seem to 
be necessary for the transmission of detail, these signals would be filtered and 
thereby limit the red bandwidth to 1 megacycle and the blue bandwidth to 
0.2 megacycles. 

The green bandwidth, being retained to the full 4 megacycles, is considered 
to be the dominant signal, and it is used to modulate the transmitted carrier 
using the ordinary vestigial sideband type of transmission. A red subcarrier 
is then chosen having a frequency such that it will fall midway between two 
of the sidebands, produced by the green signal. This is accomplished by 
separating this subcarrier from the green carrier by an odd multiple of one- 
half the line frequency. One frequency suggested by GE is 3,189,375 cycles/ 
sec which is the 405th multiple of 7,875 cycles/sec. This subcarrier is modu- 
lated with video signals from the red channel using a vestigial sideband type 
of modulation. The red sidebands would then theoretically be interleaved 
between the green sidebands with very little cross talk between them. A 
similar method would be used to multiplex the blue subcarrier on the green 
signal and the blue sidebands would be interleaved between the green in the 
same manner. 


3 a 


1 ", 3 






JJ t J 

/ \ u\ 

/V C+3.l8m.c. rCt3.89m.c. 

The optimum position of the blue carrier has not been determined. How- 
ever, one possible arrangement is shown in Fig. 8 which shows the position of 
the green carrier, the red and blue subcarriers, and the spectral region occu- 
pied by the sidebands of each color. 

In the color receiver the signal is passed through suitable filters which 
remove the portions of the spectrum which are not used by the color asso- 
ciated with its particular channel. No attempt is made to remove the inter- 
leaved information within this band of frequencies. Mathematical analysis 
shows that the cross talk which is present due to the presence of these unde- 
sired sidebands is 180 out of phase on successive frames so that the net signal 
over two frames is zero. To the extent to which the persistence of the phos- 
phors and of the eye combines the images of two successive frames, this flicker 
will not be noticed. The signals accompanying each carrier would be demodu- 
lated and would be used to actuate a direct view color tube. 


This system is compatible in that the same line field and interlace standards 
are used as for standard monochrome television. On a black-and-white 
receiver the green signal would be reproduced. According to GE "Cross 
talk [from the red and blue sidebands] would cause no trouble because it is 
geometrically in the same position on the screen as the green signal itself." 
The red and blue subcarriers would produce a dot pattern which according to 
GE has been tested and found to be unobjectionable. 

The principal advantages claimed for the GE system are compatibility, 
absence of twinkle, crawl or flicker, and the absence of any precision timing 
equipment in the receiver. The most serious objection which has been made 
to the system is the degradation of the picture due to differences in propaga- 
tion at the three carrier frequencies. The seriousness of this problem can be 
determined only by field testing. 

The Uniplex system described by CT1 employs a color switching rate be- 
tween the dot sequential system of RCA and the line sequential previously 
advocated by this company. The color repetition rate is 1.008 megacycles 
which is the 64th harmonic of the line repetition frequency. This is slightly 
more than one-quarter of the rate at which the colors are switched in the 
RCA system. Thus the color segments on each line are about four times as 
long as with the RCA system and it could perhaps be described as a "dash 
sequential" system. In the transmitted waveform, red, green and blue 
signals are sent in sequence, the color cycle being completed in approximately 
1 /z sec. Mathematical analysis shows that such a waveform will contain 
very little energy at the third harmonic of 1.008 megacycles but the second 
and fourth harmonics will be transmitted and will enable reasonably rapid 
transition from one signal level to another as would be required by varying 
amounts of primary color in the portion of the picture being transmitted. 

An ingenious color camera has been invented by CTI for use in connection 
with this system. This camera -provides the video signal with one image 
orthicon using a filter printed on 35-mm motion picture film which passes in 
front of the lens and provides the suitable color separation. The system, 
however, is not limited to the use of this camera, and the RCA three-tube 
camera could be adapted to it with ease. 

At the receiver the color switching is accomplished by a suitable gating 
circuit which is synchronized with the color switching at the transmitter by 
means of low-amplitude 3.024-megacycle signal which is transmitted con- 
tinually. This signal is 180 out of phase on successive frames and to an 
extent is canceled out by the persistence of the phosphor and the eye. Color 
phasing is accomplished by transmitting a burst of energy at 1.008 mega- 
cycles during a portion of the vertical blanking period. The demodulated 
signal is gated in synchronism with the transmitter for separation of the three 
colors and the three signals are then applied to the picture reproducing device 
whatever it may be. It is felt that in all probability this will be a direct- 
view color tube. 

The principal advantages claimed for this system are compatibility, a 
minimum amount of color contamination due to color cross talk, and con- 
siderably simplified apparatus. Simulated tests are said to have shown that 
small area flicker with this system would be no more serious than the RCA 
system even though the dots are somewhat longer. 

Color Cathode- Ray Tube 
With Three Phosphor Bands 


SUMMARY: A cathode-ray tube with a screen consisting of three phosphor 
bands fluorescing in the primary colors of red, blue and green may, in prin- 
ciple, be used to effect wholly electronic reproduction of television pictures 
in natural color if field- or line-sequential transmitting norms are employed. 
However, this tube inherently suffers from two limitations: the screen 
area is inefficiently utilized from the standpoint of light output, and the 
resolution capabilities are inadequate. 

OLOR TELEVISION PICTURES transmitted by field- or line-sequential 
\J( systems can be reproduced by a cathode-ray tube in a way which 
somewhat resembles the well-known Thomascolor system of color 
cinematography. 1 In the latter a scene is recorded on different parts 
of a film through primary color filters, for example, Wratten filters 
Nos. 26, 47 and 58. If the resulting black-and-white film images 
are projected through the same filters by a parallax-free optical system 
and superimposed, a picture in natural colors results. In a color 
television system, a corresponding series of black-and-white images, 
containing the light and shade values for the three primary colors, 
appear on different parts of the screen of a cathode-ray tube, having 
only one gun and one deflection system. 2 Primary color filters may 
be placed in front of the image sections of the tube and the color 
images derived therefrom are superimposed on a viewing screen by a 
suitable projection-optical system. Scanning of the three image areas 
of the tube screen may be in accordance with any of several well- 
understood processes. For example, with field-sequential transmis- 
sion the individual areas are completely scanned in sequence and it is 
convenient to arrange the image areas one below the other. With 
line-sequential transmission, individual lines of the image areas are 
scanned in a repeating sequence and it is preferable to have these 
areas in side-by-side relation. The choice of the image-area orienta- 
tion, the number of fields per second, the manner of interlace, etc., are 
dictated by the specifications of the color television system itself and 
will not be considered further. It is sufficient to note that in any case 
the color system is wholly electronic without any mechanical moving 
parts. Of course, since primary color images are superposed optically, 

PRESENTED: April 25, 1950, at the SMPTE Convention in Chicago. 


368 CONSTANTIN S. SzEGHo October 

it is strictly a projection system and is not applicable to direct-view 
television. Moreover, as the picture tube has only one gun, it is not 
suitable for use with simultaneous color transmissions and the pro- 
vision of discretely different image areas further limits the practical 
utility of the tube to systems of the field- and line-sequential scanning 
type because, in order to reproduce images transmitted by a dot se- 
quential system, the spot would have to fly between picture points 
located in each of the three separate image areas which is not prac- 
tically realizable. Other limitations, such as stringent requirements 
for linear scanning to insure registry of the three images, and matters 
of field or line flicker and line crawl, and the associated issue of com- 
patibility with black-and-white transmissions will not be dealt with. 
The remainder of this paper is concerned only with the reproducing 
cathode-ray tube itself. 

Fluorescent Screen 

The Wratten filters mentioned above have an average light trans- 
mission of only about 15%, so that approximately 85% of the light is 
wasted. As the color system is of the projection type which inevi- 
tably involves a considerable light loss in the optical system used for 
superimposing the primary color images, the additional absorption in 
the filters is a material drawback. Therefore, the first step in the 
development of a new color tube for use in the system was to dispense 
with these filters. A cathode-ray tube was constructed having three 
image areas capable of fluorescing in the three primary colors in 
response to electron bombardment. Specifically, the screen consisted 
of three phosphor bands: a blue band of zinc sulfide activated with 
silver, a green band of zinc orthosilicate activated with manganese 
and a red fluorescing band of zinc cadmium sulfide activated with 
silver. The spectral distribution curves of these phosphors, as manu- 
factured by Patterson Screen Division of E. I. du Pont de Nemours 
& Co., are shown in Fig. 1. In making the measurements for these 
curves, the phosphors were excited by ultraviolet light instead of 
cathode rays, but it is felt that this did not change their spectral 
distribution materially. Trichromatic coefficients (x and y), domi- 
nant wavelength and purity are given in Table I. Characteristic 
curves have been given for two red powders which were used in dif- 
ferent tubes, and it may be seen from Fig. 1, where the cutoff of 
Wratten filter No. 26 is also shown, that a substantial portion of the 
energy of these phosphors falls in the unwanted orange region. It is 
possible that cadmium phosphates or borates would have more suit- 
able spectral distribution, but they were not tried. 




TABLE I. Characteristics of the Phosphor* 

No. (Patterson) 











3650 A 

3650 A 

3650 A 

2537 A 











Dominant wavelength, m/u, 





% Purity 






608 Q-36-1080 Q-37-1243 


5 60 



^ 40 


5000 6000 



Fig. 1. Spectral energy distribution of the phosphors. 

Experimental tubes with 7- and 10-in. diameter screens were made, 
the middle phosphor bands being approximately 1% in. and 2% in. 
wide, respectively. Two methods of screen application were tried 
and they will be described with reference to the apparatus shown in 
Fig. 2. In the first method, illustrated by the diagrams at the right of 
the figure, the cylindrical part of the tube envelope is cut open ap- 
proximately 4 in. from the screen and a mask, comprising one or more 
stainless-steel sheet segments, is placed on the glass to cover approxi- 
mately two-thirds of the screen area. The blue band is now settled 
by the customary settling technique on the remaining and exposed one- 
third area of the screen. Thereafter, the settling liquid is siphoned 
out, and the screen section is baked. The mask may then be inverted, 
covering the screen section already deposited, and exposing another 
section on which another phosphor material is settled and dried. In 
the next step, the mask is arranged to cover the two coated screen 
sections, leaving the third to receive a third phosphor in a similar 
manner. This is a laborious process, and the repeated moistening 
and drying steps may reduce the efficiency of the fluorescent bands. 




In the second method employed, a settling chamber having three 
compartments made watertight with latex rubber, was used. If the 
same hydrostatic pressure is maintained in each compartment, the 
three screens may be settled simultaneously without seepage, and 
only one drying process is necessary. After the three-section fluores- 





Fig. 2. Masks used in the preparation of the fluorescent screen. 

Fig. 3. View of 7-in. cathode-ray tube with three phosphor bands. 




cent screen has been fabricated by whatever method chosen, the 
screen-bearing portion of the tube envelope is rejoined to the rest of 
the glass bulb, and the screen' is aluminized, using the customary or- 
ganic film basing and aluminum evaporation techniques. This alumi- 
num backing removes the sticking potential that may otherwise be 
encountered in the operation of the tube and nearly doubles the bright- 
ness by functioning as a reflecting mirror. Figure 3 is a photograph of 
the completed screen, showing an approximately JfJ-in. wide gap 
between the color bands which is permissible as the beam is blanked 
out when scanning of these gaps would occur. 

Consideration will now be given to the light output of such a 
banded screen. The area of one color image is approximately one- 



V) FOCUS eo .4-3 /.3 

/ r J-00>/<7 CANDLES 

I I T rr 

Fig. 4. Beam current 
density saturation of 
blue phosphors; alumi- 
num backed, 25-kv, 
525 lines, 30 frames. 
Upper abscissa scale : 
beam current; lower 
abscissa scale: raster 
current density. 

100 300 ZOO 400 SOO 600 7OO 80O 300 



5.-^ ^.5 5:2 

ninth of the entire raster area of the tube. This means that only one- 
ninth of the light flux which could be made available, is utilized. 
The situation is even less favorable when saturation phenomena of the 
phosphors are taken into account. It is found that the zinc sulfide 
and zinc cadmium sulfide types of phosphors saturate considerably at 
the high current densities prescribed by the spot size necessitated by 
the small area of one color image. Their luminous efficiency drops at 
high beam current densities, as shown on Fig. 4 where the ratio of the 
luminous efficiency of a focused raster relative to a de-focused raster 
of a blue zinc sulfide is plotted in terms of the beam current. The 
luminous efficiency 17 was measured in candles per watt. The cross- 
sectional area of the de-focused spot was twice that of the focused 
spot and the focused spot-current density measured at 400 /za 




(microamperes) had the high value of approximately J/ amp/sq cm. 
The corresponding raster-current density was relatively low as may 
be seen by reference to the lower abscissa scale of the figure. It is 
true that a nonsaturating blue phosphor, a calcium magnesium sili- 
cate, is available and its saturation characteristic is also shown on the 
figure, but this material is not very efficient, having a luminous effi- 
ciency at high-current densities of only about one-third that of the zinc 
sulfide. Consequently, its use would not improve light output. The 
green phosphor is also a silicate and saturates, although to a lesser 
extent than the blue and red phosphors used in making the tube. 

In view of the small size of each color image the raster-current 
density of the tube is unusually large and, since the heat cannot be 


100 200 300 400 500 /"<* 

! 2 3 *' 


Fig. 5. Luminous efficiency of 
blue ZnS at two different raster 
current densities; aluminum 
backed, 27-kv, 525 lines, 30 
frames. Upper abscissa scale: 
beam current; lower abscissa 
raster current density. 

18 24 


adequately dissipated, the screen heats up. This is most undesirable 
because certain zinc sulfides lose luminous efficiency at elevated 
temperatures. This is shown in Fig. 5, where the luminous effi- 
ciency for a scanned area of 5^ X 4^ in. is compared with that ob- 
tained in scanning an area nine times smaller. Of course, for sequen- 
tial systems having lower field-scan rates per second, the drop would 
not be as pronounced because the temperature rise would be less. 

Again, as a direct consequence of the high raster density or high 
screen loading, the fluorescent powders darken, an effect known as 
electron burning. When this occurs, the screen brightness drops 
15-40% in a few hours of operation, and then remains essentially 
constant at that level. Another cause for diminished brightness is 
discoloration of the screen supporting glass resulting from X rays 


released at 30 kv, which is the working voltage of this tube. Loss of 
brightness attributable to discoloration of the glass may be mini- 
mized by use of the new, nonburning glass No. 3459 of the Pittsburgh 
Plate Glass Co., instead of pyrex glass from which the first tubes were 

Surface brightness of the green phosphor was measured after a few 
hours of operation and found to be approximately 7000 f t-L at 400-jua 
beam current and 1% X 1% in. raster, 525 lines, 30 frames. Corre- 
sponding figures for the blue and red phosphors were approximately 
1700 ft-L and 1400 ft-L, respectively. With this brightness of the 
primary image areas, the highlight brightness of a colored picture on 
a 12 X 16 in. uniformly diffusing screen is in the order of 4.5 ft-L, 
if an optical system having a light-gathering efficiency of 10% is as- 
sumed and the Color Television Inc. transmitting norm is used. 

While saturation limitations may one day be overcome by new and 
better phosphors, improvement of light loss due to insufficient utiliza- 
tion of the screen area may be visualized by scanning primary areas 
that are larger and give more light even though their aspect ratio is 
incorrect, so long as appropriate cylindrical lens elements are also 
used to gather light from the whole of the scanned area and project an 
image having the correct aspect ratio. 


The postage-stamp size of each color image imposes exacting resolu- 
tion requirements, which are inherently impossible to meet with the 
present three-band tube. Approximately 475 lines must be resolved 
in each picture, a resolution of 12 lines per millimeter. Moreover, 
while one color-image area is approximately in the center of the screen, 
the other two are toward the edges, and excessive deflection de-focus- 
ing cannot be tolerated. Unfortunately, a small spot size in the 
center and minimum de-focusing toward the edges of a screen are con- 
tradicting requirements from the point of view of tube design. 

A satisfactory prototype for a projection tube which has the re- 
quired resolution on a small screen, corresponding to the central color 
area of the 3-band tube, does exist and has been described, for exam- 
ple, by H. Rinia, J. deGier and P. M. Van Alphen. 3 The outline of 
such a tube is drawn at the top of Fig. 6, which also shows below 
schematically the geometry of the scaled-up models for two 3-band 
tubes having screen diameters of 7 in. 

In the three color-band tube under consideration it is necessary to 
scan a distance in either the line or field direction equal to three times 
the corresponding dimension of one color area. In order to scan all 




three primary-image areas, two possibilities suggest themselves if it is 
required to maintain the same deflection angle a which was required, 
because it was desired to operate this tube with deflection equipment 
available for 5TP4 projection tubes having a deflection angle of 50. 
Using the notations of Fig. 6, the Helmholtz-Lagrange law defines the 
spot size in the center of the screen as follows : 

2/2(center) = ^\^Vl 

where ^(center) is the radius of spot at the center of screen; 

$ semi-divergence-angle of the beam at the cathode side; 

semi-convergence-angle of the beam at the screen side; 

EI emission energy in electron volts ; 

Ez anode voltage; and 

2/i radius of beam at crossover point. 

Fig. 6. Schematic of 7-in., three phosphor band projection tube. 

If it is assumed that the electrical conditions for the tube are to remain 
constant, the center spot size varies inversely with the angle : 




The formula expressing increase of spot radius in terms of beam 
deflection is 4 : 


2/2 (margin) = 

where 3/2 (margin) is the spot radius near the edge of the screen; 
Y half of the total deflection on the screen; 
r l beam radius in the lens plane which is approximately equal 

to the beam radius in the deflection field; 
I length of deflection field; and 
L distance from the end of deflection field to screen. 

From the geometry of the arrangement, it is seen that 

tan a 

where a is the deflection angle. 
For constant deflection angle a: 

2/2(mar K in) ^ 7*j . 

Table II gives the beam radius at the lens and the convergence 
angle at the screen for the three cases illustrated in the figure and in 
the order recited, namely, the prototype tube with but a single small 
screen, the first mentioned scaled-up model and the last mentioned 
scaled-up model. 

Case Beam radius at lens Convergence angle at screen 

IT "~ 


* e 

*l 07 1 T ' 

ol + L 
o o r 3r z 

I + L ' 

q r r i 

* l I + L 


In Case 2 the center-spot size, determined by the reciprocal of the 
convergence angle, remains small as desired because 6 remains ap- 
proximately the same as in the prototype, but the marginal-spot size, 
determined by the beam radius at the lens, becomes larger because 
this radius increases threefold. In case 3 the convergence angle be- 
comes too small and the center-spot size increases undesirably even 
though the marginal-spot size is maintained at substantially the 
desired value. It is manifest that at best a compromise solution 


must be accepted for it is not apparently possible to scan all three 
primary-image areas while preserving the resolution at both the 
center and marginal portions of the screen. 

Suitability for Theater Television 

The question which must be in the minds of motion picture engi- 
neers whether or not this tube may ultimately be used for color 
television projection in theater installations has been answered in 
the foregoing analysis. The light requirements for large screen 
work are increased manifoldly, at least 100 times for a 15 X 20 ft 
picture. The poor utilization of the screen area for light output rules 
out the three-band cathode-ray tube for this service. For mono- 
chrome theater television, a cathode-ray tube having a minimum of 
10-in. diameter and a fast Schniidt optical system are presently 
required, and it is the opinion of the author that if the complication 
of three optical systems to superimpose primary-color images is to be 
added, it would be more advantageous to make use of three cathode- 
ray tubes and simultaneous color transmissions. Local conversion 
of the sequential information for simultaneous modulation of three 
guns with the aid of storage tubes could also be visualized. This would 
increase light output by a factor of 27 and be of material help in over- 
coming the greatest hurdle of theatre projection: insufficient light 

Acknowledgment: The author is greatly indebted to Dr. E. Meschter of the 
Patterson Screen Div., E. I. du Pont de Nemours & Co., for the spectral dis- 
tribution measurements. 


1. L. R. Kistler, "The projection of Thomascolor motion pictures," Intern. Proj., 

vol. 20, no. 7, pp. 12-14, July 1945. 

2. R. Lorenzen, U.S. Pat. 2,200,285 (1940). 

3. H. Rinia, J. deGier and P. M. van Alphen, "Home projection television," 

Proc. I.R.E., Pt. I, vol. 36, no. 3, pp. 395-400, Mar. 1948. 

4. A. Wallraff, "Spot distortions in cathode-ray tubes with wider deflection 

angles," Arch, fur Elektrotechnik, vol. 29, pp. 351-355, 1935. 

A Magnetic 

Record- Reproduce Head 




SUMMARY: It is the purpose of this paper to discuss various features of a 
combination magnetic record-reproduce head of the ring-shaped type, known 
as MI- 10794, with emphasis more on the general principles of construction 
than details of assembly. Shown in the paper are various lamination shapes 
for ring-type heads; the change of head inductance with change of front-gap 
and back-gap spacer thickness; the lamination stacking factor as a function 
of lamination thickness; the shape of the lamination employed for the subject 
head; flux distribution patterns about the front-gap of a record head for 
various thicknesses of front-gap spacer; the "gap effect" of a reproduce 
head; the output voltage and inductance of a reproduce head as a function 
of the front-gap reluctance; and a photograph of the finished record-repro- 
duce head. 

RING-SHAPED magnetic record and reproduce heads were first 
described in 1935 in a paper by E. Schuller. 1 The term ring- 
shaped heads does not refer specifically to circular cores, but in- 
cludes rectangular, semicircular, diamond, trapezoidal, and even 
triangular shapes, as shown in Fig. 1. In general, therefore, a ring- 
shaped magnetic record or reproduce head may be defined as one in 
which the magnetic material forms a quasi-toroidal enclosure with 
one or more air gaps, and with the magnetic medium bridging one 
of these gaps, commonly spoken of as the front-gap, and contacting 
the structure on one side only. 

When a second gap is inserted in the core so as to divide it into two 
symmetrical halves, the second interstice may be termed the back- 
gap. Each hiatus usually contains a nonmagnetic spacer, the 
"front" and "back" spacer, although some commercial ringheads 
have butt joints. One or more coils may be wound around the core, 
and the entire assembly is usually placed in a high-permeability can 
to act as a magnetic shield. It is the purpose of the following to dis- 
cuss various features of one such head, known in RCA as the MI- 
10794 combination magnetic record-reproduce head, with emphasis 
more on general principles of construction than details of assembly. 

The inductance of the head varies with the thickness of both the 

PRESENTED: April 24, 1950, at the SMPTE Convention in Chicago. 





Fig. 1. Core shapes 
for magnetic heads. 

front- and the back-gap spacers, and is given approximately by the 

KN 2 

Ju - 

r + R 
where r reluctance of front-gap; 


I gap-length, cm, 
d = gap-width, cm, 
w = gap-depth, cm, 
R = reluctance of back-gap; 

w d 

1 = gap-length, cm, 

w = gap-width, cm, 

d = gap-depth, cm, 

N = number of turns, 

K = constant. 

The above equation is true as long as the reluctance of the core 
material is small compared to the reluctances of the gaps; otherwise, 
an additional term for the core reluctance must be introduced in 
the denominator of the above equation. Figure 2 shows the varia- 
tion of inductance with thickness of front- and back-gap spacers; 
these curves, calculated by the above equation, correspond closely 
with observed values. 

Magnetic record heads of the ring-shaped core type are usually 
constructed with a low-impedance winding, and reproduce heads, 
with a high-impedance winding. The reason for the low impedance 
of a record head lies in the fact that the line to the head has a certain 
amount of capacity, which represents a leakage path for the high- 
frequency-bias current which is almost always used. If the record 
head had a high inductance, the line leakage would constitute a con- 




siderable loss. In the case of the reproduce head, where no bias fre- 
quency is employed, a high-impedance winding on the head is satis- 
factory, provided the length of the cable connecting the head with 
the reproduce preamplifier is short and otherwise is so constructed 
as to have a low leakage capacity, otherwise high-frequency signal 
losses may be suffered. 

L O =.OIO" 


DO -.100" 

to W -W-.200" 


.1 .2 .3 .4 .5 .6 .7 .8 .9 1.0 I.I 1.2 1.3 1.4 1.5 1.6 1.7 1.8 


L-.0002" D =.100" D=.OI5 M W =W=.200" 










1 . 

10 20 30 


Fig. 2. Upper part of figure shows variation of head inductance with thickness of 
front-gap spacer. Lower part of figure shows variation of head inductance with 
thickness of back-gap spacer. 

For the head under consideration, which can be used either as a 
record or as a reproduce head, the inductance is 5 mh (millihenrys) . 
This means, essentially, that a step-up transformer with a high 
turns ratio is required in the associated amplifier when the unit is 
employed as a reproduce head. As such, the combination is equiva- 
lent to a reproduce head with a 2-h (henry) inductance, which, of 
course, represents a head of impractical construction. Therefore, 




since for maximum voltage on the first tube grid of the reproducing 
amplifier a step-up transformer is still necessary with heads of feasible 
or practicable inductances, a reproduce head with a 5-mh inductance 
does not appear to present difficulties or complications. 

Besides the number of turns on the winding and the thicknesses 
of the front- and back-gap spacers, the inductance of the head is 
determined by the lamination material and the number of lamina- 
tions in the core. Extremely thin laminations, say 2 mils thick, to 
reduce eddy current losses to a minimum, are difficult to handle and 
exhibit a low stacking factor. By stacking factor is meant the per 
cent space occupied by the laminations themselves, exclusive of the 
space taken up by the cement between the laminations. As an ex- 


Fig. 3. Variation of space or 
stacking factor of magnetic head 
core with lamination thickness. 

I 2 3 4 5 6 7 8 9 10 II 12 

ample, consider a core 200 mils wide, containing 80 two-mil lamina- 
tions held together by 79 layers of cement each 0.5 mil thick. In this 
case the space factor would be 80%. When 6-mil laminations are 
used, held together by layers of cement each 0.5 mil thick, the actual 
space taken up by the laminations in the 200-mil-wide core comes to 
92%. Figure 3 shows the space factor as a function of lamination 
thickness for a 200-mil core. Thicker laminations mean not only 
that fewer turns of wire will be required to obtain the desired in- 
ductance, but also that the flux density in the laminations will be 
less, because the space factor is greater, thereby reducing possible 
distortion in the head at high audio and bias currents. 

Figure 4 shows the approximate shape of the lamination used. 




In the case of the record head, the front-gap exists for the purpose 
of providing a leakage flux path for magnetizing the magnetic tape 
as it passes over the head in recording. Concerning the reproduce 
head, if no front-gap reluctance existed, the flux from the tape 
would merely pass through the pole tips and not through that part 
of the core on which the coil is wound and in which it is desired to 
induce an electromotive force. 

Fig. 4. Ap- 
proximate shape 
of lamination 
used in subject 
magnetic head. 














504030 20 10 10 20 3040 506 


Fig. 5. Leakage flux distribution about front-gap of a 
ring-type magnetic record head, in the absence of a record- 
ing medium, when spacers of different thicknesses are in- 
serted in the gap while constant current is supplied to the 

Figure 5 shows leakage path distributions about the front-gap of a 
ring-type magnetic record head, in the absence of a recording medium, 
when spacers of different thicknesses are inserted in the gap while 
constant current is supplied to the head. The spacer serves the 
double purpose of maintaining the nonmagnetic gap parallel and to 
avoid the accumulation of ferrous dirt which would change the per- 
formance of the head. The curves were obtained by moving the 
head past a single loop of No. 46 wire. The head was fastened to a 
brass rack, and the pinion for the rack was driven by the curve 

382 M. RETTINGER October 

recorder; in this manner a measure of the flux density with respect 
to the distance traversed by the head was secured on the curve paper. 
The greater leakage flux with increasing thickness of front-gap spacer, 
or increased front-gap reluctance, is clearly evident. 

It should be noted again that the curves were obtained without a 
magnetic medium over the gap. Hence, they are useful chiefly for 
comparing heads and do not necessarily show the actual leakage 
flux distribution when the head is in operation; that is, when mag- 
netic tape is lying on the head or when it is passing over it. 

In an RCA standard head, the front-gap spacer consists of a very 
hard nonmagnetic alloy. This alloy is considerably harder than the 
lamination material, and hence prevents the forming of burrs on the 
pole tips. These burrs tend to short-circuit the gap, and to alter 
the head performance. To prevent the accumulation of electrostatic 
charges on the head, the spacer is grounded to the mumetal housing, 
as is the cable shield. 

After the head is assembled, the tape-bearing surface of the cores 
must be polished to secure proper scanning. This is obtained by 
lapping the head with successively finer aluminum oxide or silicon 
carbide papers in a jig to give the required contour on the head. 
Final polishing leaves a well-defined gap, free of bridging burrs, 
when examined with a 500 X microscope. 

For obtaining sufficient high-frequency response, particularly as 
far as the reproduce head is concerned, the "length" of this gap is 
very important. As far back as 1935, E. Schuller, in the paper cited 
at the beginning of the article, noted that "the magnitude of this 
magnetic gap, that is, the extent of the recording magnetic field in 
the direction of the tape travel, should be approximately one-fifth 
of the smallest half-wave length." This criterion is more applicable 
to d-c biasing, however, than to a-c biasing, according to S. J. Begun, 2 
who writes that "when a-c bias is employed, the effect of the length 
of the recording gap is practically negligible when the recording 
field is uniform and sharply defined." Hence, by employing for the 
record head a thinner than necessary front-gap spacer, namely, the 
spacer required for the reproduce head (see below), some loss in 
sensitivity is incurred, as may be seen from Fig. 5. 

Schott 3 has determined that the output of a reproduce head varies 

. ird 
sin -r- 




where d = effective gap-length, 
X = recorded wavelength. 

The physical gap-length, however, is not equal to the effective gap- 
length, the latter being from 10 to 50% larger. Figure 6 shows the 
"gap-effect," as calculated by the above equation. 


DB = 20 LOG 

-oir =20LOG 













Fig. 6. Figure shows "gap effect" of a reproduce head, or variation 
of reproduce head output with frequency for various front-gap spacers. 


A back-gap is frequently introduced in ring magnetic record and 
reproduce heads to reduce d-c magnetization with a consequent 
lowering of the noise produced by such magnetization. This is 
effected by ' 'shearing" the hysteresis curve of the lamination ma- 
terial, as shown in Fig. 7. When the d-c magnetizing force in a 
closed ferrous ring is reduced to zero, the ring will have a residual 
induction as indicated by the point A in the figure. When a large 
back-gap is inserted in the toroid, however, a "shearing" of the 
hysteresis curve is effected, and the remanent induction, B, becomes 
rather small. 

It may now be desirable to point out significant differences between 
record heads and reproduce heads for a better understanding of these 




Fig. 7. Figure shows effect of 
"shearing" the hysteresis loop 
of the magnetic record head lam- 
ination material. 

Head resonance 


Recording field 
Magnetic material 



Record Head 

Frequency of resonance, de- 
termined by head inductance 
and leakage capacity, should 
be higher than bias-current 

Relatively unimportant, be- 
cause field generated by head 
is much greater than external 

Effective gap-length is, 
within limits, not critical. 

Should show low eddy-cur- 
rent and hysteresis losses, to 
prevent heating and mag- 
netic saturation at bias-cur- 
rent frequency. 

Should not be high, to pre- 
vent bias-current losses in 
head and leads to head by 
leakage capacity. 

Within limits, not critical. 

Reproduce Head 

Frequency of resonance 
should be higher than high- 
est frequency to be repro- 

Very important because field 
generated by recorded me- 
dium is small. 

Effective gap-length should 
be as small as possible. 

Heating and magnetic sat- 
uration are unlikely to occur. 

Not critical. 

Should be small to reduce 
hum pickup. 

The preceding is concerned chiefly with a delineation of the signifi- 
cant characteristics, both in performance and construction, of mag- 
netic record and reproduce heads. The following presents a mathe- 
matical analysis of these units, more from the point of view of design 
considerations than from the standpoint of circumscribing their 
operational behavior. 

The action of a reproduce head may be explained by means of 
Fig. 8. Assuming the tape to be a constant flux (high reluctance) 
generator providing a flux <, we have across the tape head : 




Fig. 8. Magnetic schematic of a 
reproduce head. 


Magneto-motive force = <}> R t 

= 00 


r + R 


4> = flux through the core 
r = reluctance of front-gap 
R = reluctance of rear-gap plus core reluctance 

The voltage generated in the coil of the reproduce head will be 

dt r + R 

r + R 
A measure of the reproduce head efficiency is given by* 

r + R 

r + R 

where, as before, r = front-gap reluctance 
R = back- gap reluctance 
KI, KI, K% = constants 

N = number of turns 

L = inductance of reproduce head 

E = output voltage of reproduce head 

* Advanced by L. J. Anderson, Radio Corporation of America, RCA Victor 
Div., Camden, N. J. 




If in the above equation we set R = ar, we get 

, _ K s r 
Q ~ 1 + a 

If r is held constant, the efficiency is proportional to 1/(1 + a); 
see Fig. 8A. 



10 LOG 






















- -"" 



Fig. 8A. Loss in decibels, in reproduce head sensitivity as the 
back-gap spacer is made thicker while maintaining constant head in- 
ductance and front-gap reluctance. 

The voltages of two reproduce heads of similar construction and 
equal number of turns are related as follows : 

+ Rz) 

where 7*1, r% = front-gap reluctances of heads 1 and 2 respectively, 
Ri, Rz = back-gap reluctances of heads 1 and 2 respectively, 
EI, EZ = output voltages of heads 1 and 2 respectively. 
If we now make the following substitutions 

r*i = ar 2 
RI = RZ == R 
R = br z 

the above equation can be written : db = 20 log a _^_ ^ 




It is seen (Fig. 9) that an increased reluctance at the front-gap 
(as when wear on the head reduces the pole face depth, d, in Fig. 2) 
increases the output voltage from a reproduce head. This increase 
in the voltage is to some extent dependent on the reluctance of the 
back-gap. In Fig. 9 the abscissa represents the factor by which the 
reluctance of the rear-gap is larger than that of the front-gap, and in 
which the ordinates show the gain in output in decibels resulting from 
increasing the front-gap reluctance by the factor a. For example, a 
head having a rear-gap reluctance equal to the front-gap reluctance 
(5=1) will provide a 2.6-db gain in output when its front-gap re- 




Fig. 9. Variation of magnetic reproduce head output with ratio 
of rear-gap to front-gap reluctance for two kinds of head. 

luctance is doubled (a = 2), as when wear on the head has reduced 
the pole face depth, d, to one-half its former dimension ; a head hav- 
ing a rear-gap reluctance which is ten times as great as the front-gap 
reluctance (b = 10) will provide a 5.3-db gain in output when its 
front-gap reluctance is doubled (a 2), as when wear on the head 
has reduced the pole face depth, d, to one-half its former dimension. 
Whenever wear on top of the head increases the front-gap reluc- 
tance, a decrease in the inductance occurs. This variation in in- 
ductance is given by 

L 1 

r 1 + R 
r + R 




where L = inductance of head when head was first put in operation, 
L 1 = inductance of head after wear on top of head has in- 
creased its front-gap reluctance, 

r = front-gap reluctance when head was first put in operation, 
r 1 = front-gap reluctance after wear on top of head has de- 
creased the pole face depth, d, 
R = rear-gap reluctance. 

In Fig. 10, the abscissa represents the pole face depth in mils and 
the ordinates indicate the inductance and the gain in output voltage 
in decibels from a reproduce head as the top of the head becomes worn. 





9 2 


7 S 









































16 18 

2 4 6 8 10 12 


Fig. 10. Variation of reproduce head inductance and reproduce 
head output with pole face depth of front-gap. 

Assuming the initial pole face depth to be 15 mils, then the inductance 
of the head is 5 mh. When wear on the head has reduced the depth 
to 5 mils, the inductance has decreased to 4 mh while the output 
voltage from the head has increased 7.5 db. If the front-gap is 
truly a rectangular parallelepiped, so that the gap does not become 
wider with decreasing pole face depth, the frequency response will 
remain relatively unaffected even though the sensitivity increases. 


The core assembly is embedded in plastic which, at the same time, 
acts as the bonding agent between the two telescoping high-per- 
meability cans which comprise the housing for the head. The plastic 




has a dielectric constant of 3.7 at 50 cycles and 3.6 at 1 megacycle, at 
22 C, which is practically equivalent to that of quartz. 

The fact that the cores are not clamped mechanically, always an 
undesirable condition for permalloy laminated structures, but are 
embedded in a substance of greatly different mechanical impedance 
than the metal housing, makes the head singularly free of micro- 

Fig. 11. Photograph of MI-10795 magnetic record-reproduce head. 

CD -|0 






10,000 20,000 


Fig. 12. Combination of subject head response and magnetic film characteristic 
for two types of film when constant current is supplied to the record head and the 
recording is reproduced on the same head. 

Figure 11 shows a photograph of the completed head. The 18-8 
stainless steel stud, shot-welded to one of the mumetal cans, is used 
for fastening the unit to the ball of a ball-and-socket type of mounting 
which allows longitudinal, lateral and transverse adjustments 4 of the 

Figure 12 shows the combination of head response and film char- 
acteristic for magnetic film A and magnetic film B, when constant 


current is supplied to the record head and the recording is reproduced 
on the same head and measured with a high-impedance vacuum tube 
voltmeter. The 68-kc bias current for the measurement was .016 
ma; the signal current, 2 ma; the film speed, 18 in. /sec; and the 
output voltage (in the case of the A film) at the peak frequency for 
100% modulated track,* 2 mv. Since demagnetization effects on the 
film are not readily dissociable from gap effects, Fig. 12 does not 
represent the frequency response of the head; it does still, however, 
serve as a valuable criterion for the head response to those familiar 
or experienced with the subject. 

The head is % in. in diameter, and thin enough so that three heads, 
without studs, can be mounted in a line to produce, on 35-mm mag- 
netic film, a triple track with each .200 in. wide. 


Summ arizing, this head has proven to be a good commercial product 
for the?* reasons : 

1 . High quality performance : This includes not only the extended 
frequency-response range from 30 to 18,000 cycles, but also high 
sensitivity, absence of microphonics, low bias-current requirements 
and low hum level due to its small size (approximately 3 cu/cm or 
.183cu in. /cm). 

2. Low manufacturing cost in respect to quality : This is achieved 
by the manufacture of a larger number of identical parts, with a 
consequent reduction of cost per item, and by the reduction of the 
actual number of components in each head, particularly machined 
parts, such as mechanical clamps, terminal blocks, etc. 

3. The reproduce head is identical with the record head, so that 
the user requires fewer spare parts for his recording machine. 

Acknowledgment: The author is grateful to Mrs. E. Addington for 
valuable assistance rendered in the construction of the subject head. 


1. E. Schuller, "Magnetische Schallaufzeichnung," Electrotechnische Zeit- 

schrift, pp. 1219-1221, Nov. 7, 1935. 
See also German AEG patent No. 617796 (1935). 

2. S. J. Begun, Magnetic Recording, p. 64, Murray Hill Books, Inc., New York, 


3. WienSchott, "Einfluss der Schragstellung des Spaltes bei Intensitatschrift," 

Zeitschriftder Techn. Physik, vol. 17, p. 275, 1936. 

4. Terms are those defined by N. M. Haynes, "Magnetic tape and head align- 
ment nomenclature," Audio Engineering, vol. 33, no. 6, p. 22, June 1949. 

* 100% was defined as the recording level at which 2% second harmonic distor- 
tion is obtained at 400 cycles. 

Physical Principles, Design and 
Performance of the Ventarc 
High- Intensity Projection Lamps 



SUMMARY: The Ventarc lamps represent a new series of high-intensity 
carbon arc lamps for motion picture projection recently developed by the 
author at Zurich, Switzerland. The blown arc of the new lamp produces a 
distribution of brilliance which is highly advantageous for the illumination 
of the aperture. By using a new negative electrode the arc can be pushed 
up to an extremely high brilliance. The precision feed control of the Ventarc 
insures a perfect homogeneity and invariability of the screen illumination. 
The visible radiation of the arc is effectively concentrated on the projecture 
aperture by an entirely new optical system. Heating of the film is minimized 
by eliminating all invisible radiation. No surplus energy not useful at the 
screen has to pass through the film. 

ANEW SERIES of high-intensity projection arc lamps has been 
developed by the Edgar Gretener Company at Zurich, Switzer- 
land. These lamps employ a number of fundamentally novel design 
features which are of considerable technical interest and which form 
the basis for this paper. 


The light source employed in the Ventarc lamps is a " blown arc," 
invented by the author some years ago. Some historical remarks 
will be of interest before particulars of the present design are dis- 

Experiments were made in 1932 at the central laboratories of 
Siemens, Halske A.-G., Berlin, in connection with the development 
of a new super high-intensity lamp for use with lenticular color film 
projection systems. This lamp was to compensate, by increased 
crater brilliance, for at least part of the light lost in the color filters, 
and was to realize a distribution of brilliance over the positive crater 
rotationally symmetrical and constant in time. These requirements 
are essential for correct color rendition and brightness in projecting 
the Berthon-Siemens film. At that time a solution of this problem 
was found by employing air blast concentration of the arc, water- 
cooled jaws clasping the positive carbon immediately behind the 
PRESENTED: April 27, 1950, at the SMPTE Convention in Chicago. 





crater and a radical reduction in thickness of the positive carbon 

It is very interesting to see that the latest trends in development of 
normal high-intensity arcs for high light output follow fundamental 
principles already discovered in the course of the above-mentioned 
development activities, e.g., water-cooled contacts, positive carbons 
with thin shell and operation with short protrusion. In comparison 
with these, air-blast concentration of the arc has not, to date, been 
given the attention it deserves, in our opinion. 

Investigation of brilliance distribution in the positive crater of a con- 
ventional 150-amp lamp with rotating positive carbon led to the 
following conclusions (Fig. 1): 

Plans of 

y brilliancy 




Fig. 1. High-intensity arc. 

1. Taking the average in time, the brilliance distribution is sym- 
metrical with respect to that plane, through the axis of the positive 
carbon, which separates the tail flame of the arc into two identical 

2. The peak value of brilliance is located above center, in that 
region where the arc tail flame issues from the crater, i.e., where the 
product of current density multiplied by the volume of gas assumes a 
maximum value. 

The lack of rotational symmetry in the distribution of brilliance in 
a customary Beck arc is due to the fact that the current of anodic gases 
and the electrical current follow diverging paths outside the positive 
crater. Consequently, gas concentration and electric current density 
are not distributed symmetrically about the axis of the positive car- 

* E. Gretener, "A brief survey of the physics and technology of the Berthon- 
Siemens color process," Jour. SMPE, vol. 28, pp. 447-463, May 1937. 




Such observations led to the following conclusions, which were then 
made the basis for development of a new lamp (Fig. 2) : 

1. Both the current of anodic gases and the electric current have to 
be aligned with the axis of the positive carbon in order to obtain the 
desired rotational symmetry of brilliance distribution. This may be 
obtained with coaxial electrodes by directing a slightly converging 
conical air blast, along the positive carbon, which issues from the 
housing of the positive carbon and proceeds toward the negative 
electrode. This air blast produces an effective concentration of the 
anodic flame in the cylindrical space in front of the positive crater. 

2. The current load may be raised to a maximum value by employ- 
ing a positive carbon with a very thin shell. This shell is burned off 
by the oxidizing effect of the air blast, leaving only a thin crater edge, 

Air Blast 






Carbon Guide 


Fig. 2. Blown arc. 

but the insulating effect of the air blast always confines the discharge 
column to the front of the positive carbon. 

3. The positive carbon is surrounded by water-cooled jaws imme- 
diately behind the arcing end which protect the rear parts of the shell 
from oxidation by the air blast and avoid excessive calcination of the 
forward portion of the positive carbon. This measure prevents a 
decrease of the initial higher brilliance toward the lower more stable 
value which is generally observed after a new carbon is burned in. 

4. The arc plasma is put into fast rotation by means of a current 
coil coaxial with the positive carbon. This increases the symmetry of 
distribution of brilliance with respect to the carbon axis, and also 
avoids formation of dark areas in the arc plasma, which would other- 
wise appear at extremely high current densities. 

Rotation of the arc plasma is caused by the coil (Fig. 3) producing 
a magnetic field, H, which diverges from the positive toward the 




negative terminal of the arc. Contrarily, the electrical flux lines, 
j, of the arc converge from the positive crater toward the tip of the 
negative electrode. Due to the rotational symmetry of both these 
fields, a resultant torque is produced which puts the arc plasma in rapid 
rotation about the arc axis. 

As a result of all these features the blown arc possesses practically 
unlimited current capacity. Provided the thickness of carbon shell 
is suitably varied, the brilliance increases in linear relation with the 
arc current and seems to be limited only by the consumption rate of 
the positive carbon. The voltage characteristic of the arc is suffi- 
ciently positive so that this fact, together with the fixed shape of the 



Shream Lines 



H P \ X 

Fig. 3. Magnetic rotation of arc. 

arc stream, permits the use of a power supply only slightly higher in 
voltage than the arc itself. 

Practical difficulties are presented by formation of a mushroomlike 
growth of rare earth carbide on the tip of the negative electrode if the 
lamp is run with short arc gap and high consumption rate of the 
positive carbon, but such difficulties can be overcome by appropriate 
construction of the negative electrode. 

Today the blown arc of the Ventarc lamp, which has undergone 
important further development during the past years, shows the 
following special features : 

1. With increased current loading of the positive core, the maximum 
brilliance, as observed from outer mirror zones, is located between the 
center and the outer edge of the crater. A distribution of brilliance 




across the positive crater is obtained as shown in Fig. 4, which permits 
uniform illumination of the film gate with high efficiency when used 
in conjunction with the optical system newly developed for the 
Ventarc lamps. Figure 4 shows the brightness variation across hori- 
zontal and vertical crater axes, in a plane perpendicular to the car- 
bon axis, as recorded from a 50 angle of view to said axis. It is seen 
that the distribution of brilliance is very flat, thus giving a high aver- 
age brightness in comparison with the peak value. 

2. The air blast employed for arc concentration will not cause 
additional oxidation if appropriate choice is made of thickness and 

wo $00 epo ifpo 2 
















ar to 
n axis 
j A-A 


















/ Along the carbon axis 






\ s 



















Fig. 4. Blown arc: distribution of brilliancy b; a = 50. 

material of the carbon shell and of velocity of the air blast. Experi- 
ments employing nitrogen and carbon dioxide instead of air show prac- 
tically the same result as regards brilliance and consumption rate. 

3. Even without water cooling, the blown arc is stable and rela- 
tively indifferent to overload. If positive carbons of the best brand 
available today are used, no decrease of initial brilliance will be ob- 
served in the interval during which a new carbon burns in. It is, 
however, necessary that the carbon be pre-cratered in manufacture 
to the form it will assume in operation. 

4. The Ventarc lamps do not produce soot, since complete oxida- 


tion of the material evaporated from the core is effected by the con- 
tinuous supply of fresh air. 

5. Operation of the blown arc is independent of the position of the 
lamp in space, which is very convenient for projection with steep in- 
clination of the optical axis. 

6. Since the direction of air and vapor currents is symmetrical 
around the edge of the positive crater, this edge will always be formed 
perpendicular to the axis of the positive carbon. The necessity of 
rotating the carbon is thereby avoided, although this is quite indis- 
pensable in normal Beck high-intensity lamps operated with high 
current -density. 

7. As viewed from the side, concentration of the arc by the air 
blast produces a very abrupt change in brilliance at the border line be- 

Positive / \\ / \ ^ \ Negative 

Electrode f^^ )^//^\ Electrode 

arc stream 

=- Direct rays from P 

-* Reflected rays coming from P 1 

Fig. 5. Auxiliary mirror. 

tween the positive shell and the intensely luminous anodic flame. As 
this border line is always at right angles to the carbon axis, it offers 
an ideal reference point for a high precision feed control which will 
position the positive crater exactly at the focal point of the concave 


Problems posed by the negative electrode of a blown arc have 
caused considerable difficulties. It must, however, be understood 
that in arcs of high current-density different conditions prevail for 
short gaps and long gaps. 

1. For special purposes a lamp with an extremely long gap was 
developed, which produces a luminous flux of 5,000,000 1m drawing 
350 amp at 180 arc volts with a gap of 80 mm. 

By the use of a small water-cooled auxiliary mirror, an inverted 
image of the arc was focused on the arc itself (Fig. 5). Thereby the 




brilliance measured at right angles to the arc axis was raised to a 
value of at least 1100 to 1200 IC/mm 2 (International Candles per 
square millimeter) along the entire length of the arc (Fig. 6). The 
positive carbon employed was 12 mm in diameter with a core of 9 
mm; the negative carbon was 11 mm in diameter. In spite of the 
enormously high output of the arc and increased consumption rate of 
the positive carbon, the negative electrode gave no difficulty, since 
sufficient oxygen penetrated the arc to prevent formation of mush- 
rooms on the negative tip. Therefore, by operating a blown arc at the 
relatively high voltage necessary with a normal high-intensity arc with 
angular trim, the gap can be enlarged so that formation of carbide de- 
posits on the tip of the negative carbon is avoided. 

Arc Straam 

I 80mm 

Dmage of the positive carbon 
Fig. 6. Brilliancy of arc-stream at 350 amps, 180 volts. 

2. For projection lamps, arc voltage should be as low as possible 
because a long arc wastes energy with additional heating of the lamp 

Attempts were made to reduce the formation of carbides at short 
arc lengths by appropriate choice of material for the core of the nega- 
tive electrode. A basic solution to these difficulties could not be ob- 
tained in this way, however. 

Formation of mushroomlike deposits is caused by evaporation 
products which deposit on the negative tip inside the plasma of the 
arc, where they cannot be oxidized. Thus the deposit on the nega- 
tive tip continually enlarges. This difficulty is avoided if the evapo- 
ration products tending to deposit on the negative tip are transported 
instead out into the air, where they rapidly oxidize. 

This continuous transport is accomplished by replacing the cus- 
tomary rod-shaped negative electrode with a slowly rotating thin disc 
of graphite, the negative spot of the arc being located on the sharp 




edge of this disc (Fig. 7) . The gases of the arc flame are exhausted by 
two suction pipes, one on each side of the disc. Reduced thickness of 
the graphite disc, sharpening of its edge, and arranging of the disc 
in a meridianal plane of the optical system reduce additional shadow- 
ing of the luminous flux by the cathode disc to a negligible value. 

This disc-shaped cathode permits satisfactory operation of a highly 
loaded blown arc with extremely short gap. Measurements taken in 
1948 (Fig. 8) show continuous increase of the brilliance for increasing 
arc currents. The brilliance attainable in such lamps seems to be 
limited only by the consumption rate of the positive carbon. A 12- 
mm positive carbon with 9-mm core, operated at 420 amp, 75 volts 
produced 2000 c/mm 2 at 60 to the carbon axis, with a consumption 
rate of 62 mm/min. This consumption could be reduced by using di- 
rect water-cooled jaws in combination with special carbons developed 

Arc Stream 

Fig. 7. Disc lamp. 

by National Carbon. It must also be understood that this brightness 
value in no way represents a limit that may not be surpassed. 

In such a lamp an approximately linear relationship exists between 
arc current and brilliance. The positive carbon used in the present 
lamp is correctly adapted to a current of 200 amp. For higher values 
of arc current, shell thickness would have to be reduced. By appro- 
priate choice of material, carbon shell thickness and velocity of the air 
blast, even better results might be obtained. 

In these experiments, the core of the positive carbon employed was 
designed for normal searchlight use and therefore was of special com- 
position to ensure stability of the arc. Since this problem does not 
arise in the blown arc, the salts in the carbon may be chosen on the 
basis of maximum brilliance only thus yielding even more favorable 




A practical difficulty in adapting the disc arc lamp for use in cinema 
projectors is presented by the comparatively short focal length of the 
reflector. In order to ensure a sufficiently long period of operation, 
the diameter of the disc must exceed the focal length of the usual 
mirror. In order to locate the crater of a disc lamp at the focal 
point, a slot would have to be provided in the reflector to accommo- 
date the rear part of the disc. This would complicate lamp design, 
particularly for the suction pipes, which would then interfere with the 
support, the drive and the means for current supply to the cathode 

Such difficulties are avoided by employment of a ring cathode in 
place of a disc (Fig. 9). The ring passes around the positive head 
and the arcing spot is located on the sharpened inner edge of the ring. 
























ZOO 300 tOO SOO Amps 

Fig. 8. Brilliancy and consumption rate of a short blown arc. 

Appropriate design and arrangement of the lamp mechanism reduce 
the additional shadowing caused by the cathode ring to a negligible 

The active part of the cathode ring is made of artificial graphite set 
in a metal mounting. The life of a single ring in a 100-amp Ventarc 
lamp is approximately 15 hr. Easy replacement is provided. The 
drive mechanism for rotating the ring as well as the supply of current 
to the ring is effected through the metal mounting. 

The inner edge of the ring is guided by two ceramic rollers located 
on each side of the positive head, so that the length of the arc gap is 
practically independent of consumption of the inner edge. The ring 
support may, however, be adjusted manually to regulate the value of 
arc current, although no such adjustment is required during a normal 
projection period. 


Extremely high operational stability is obtained by a Ventarc lamp 
with the ring cathode, as the arcing spot on the ring is kept practically 
fixed in space. 

Striking of the arc is effected by feeding the positive carbon forward 
until it touches the cathode ring and subsequently withdrawing it to 
the normal operating position. 


1. Separate positive and negative carbon feed controls are 
employed in the Ventarc lamps. Requirements for positive feed 
control, to maintain the crater in the focal plane of the illumination 
system, increase in precision with the efficiency of the illumination 

For accurate photoelectric feed control, the shape of the arcing 
edge of the positive crater must remain constant and in a plane at 

negative ring 

Fig. 9. Lamp with negative ring-electrode. 

right angles to the axis of the carbon, a condition readily met by the 
air-blown arc. 

In order to obtain maximum sensitivity of the regulating system, 
it is desirable to increase, as far as possible, the sharp change in brilli- 
ance between the forward edge of the positive shell and the base of the 
anodic flame. 

A blown arc is capable of meeting these requirements as extremely 
high brilliance is obtained in front of the positive carbon by the con- 
centration of the anode flame. This difference in intensity between 
the anode flame and the positive carbon is greatest in the blue and 
ultraviolet spectral region. 

With these favorable features a blown arc permits the use of a very 
simple and accurate positive feed control (Fig. 10). This method 
may be employed without difficulty, even in blown-arc lamps with are 
currents as low as 25 amp. By means of a small lens the crater edge is 
imaged at right angles to the carbon axis upon a slotted diaphragm. 


Behind the slot is a phototube (RCA 929) which is particularly sensi- 
tive to blue and ultraviolet radiation. This phototube governs the 
feed of the positive carbon by a glow discharge tube and magnetic 
clutch. The feed mechanism is actuated as soon as the carbon burns 
back beyond its normal position, and when the crater protrudes a bit 
too far, the feed control is stopped. Such a control is accurate within 
==0.1 mm and is practically free of inertia so that it may execute sev- 
eral feeding cycles per second. As the regulation system does not 
contain electrical contacts, the operational reliability is very high, 
with practically no breakdowns. Any manual misplacement of the 
positive carbon is automatically corrected. 

2. In the smaller Ventarc lamps equipped with a conventional 
negative carbon, the negative feed control is based on keeping the arc 

Clutch Control 

Photo Cell 

1 ] _J I Diaphragm 


PosFeed \|U Carbon 

Lens + Mirror 

Fig. 10. Positive regulation. 

current constant. If current is low, the negative carbon is fed for- 
ward, shortening the arc gap until the current increases to its stand- 
ard value. If the current is high, the reverse effect is obtained 
by increasing the length of the arc. 

This method of feed control reaches accuracies of =*= 2% by an ex- 
tremely simple design without electrical contacts and is consequently 
of very high reliability. 

In large Ventarc lamps equipped with disc or ring cathodes, the 
cathode is manually adjusted so that the gap corresponding to the de- 
sired standard value of the arc current is obtained. Due to the ex- 
tremely low consumption rate of the cathode, resetting of the 
cathode is not necessary during a projection period. 

3. Because of the high stability and constant brilliance in a blown 
arc and because of the favorable conditions for positive and negative 
electrode feed control, a steadiness of screen illumination is obtained 




amounting to ==2% to 3% over a whole projection period. More- 
over, this is accomplished automatically and independently of the 
attention of the operator. 

Life tests of the positive feed control, extending over more than 
10,000 hr, show excellent results. Two Ventarc lamps in commercial 
use in a motion picture house at Zurich throughout the past year 
have required no readjustment of the positive feed control up to the 
present time. 

A Tilting Angle 

Fig. 11. Generation of new mirror. 

1. Focal Plane 2. Focal Plane 

Fig. 12. New mirror for Ventarc lamps, four-focii type. 


1. The Ventarc lamps are equipped with a new optical system which 
is adapted to the particular distribution of brilliance of the blown arc 
and therefore gives excellent uniformity of screen brightness with high 
light efficiency. Referring to Fig. 11 the reflecting surface of the 
mirror is generated by rotating round an axis the arc of an ellipse the 
main geometrical axis of which is inclined to the rotation axis by an 
angle, 6. The value of this tilting angle, 5, is determined by the 
desired location of the crater images on the aperture plate and the 
focal distances involved (Fig. 12). As the axis of the generating 




ellipse is tilted in such a way that its two focal points do not lie on 
the optical axis, the reflector possesses two focal circles, oriented at 
right angles to the optical axis of the projection system. One of 
these circles is made to coincide with the crater edge and the second 
one is made to circumscribe the aperture (Fig. 12) . 

Ideal properties of such an optical illumination system are obtained 

1 Center 

2 Side 

3 Corner 


32 1 23 

Fig. 13. Light distribution through the film aperture. 

32 23 

Fig. 14. Light distribution through the film aperture. 

if the crater diameter, D c and the axial magnification ratio of the 
mirror, M at are made to satisfy the following condition : 

in which D g is the diameter of the focal circle at the aperture, and the 
constant a has a value of unity. 

In this case the light distribution on the aperture is as shown by 
curve I, Fig. 13. 

With the same focal circle, D g , maintained at the aperture, but with 
D c M a made greater than D 0) i.e., a > 1.0, diverging distribution curves 
are obtained as additionally shown in Fig. 13. The greater extent 
of these latter curves is the result of crater images tangent inside the 


focal circle, but larger than the aperture diagonal and so indicative of 
reduced efficiency. 

If the product, D c -M a , is smaller than, D g , i.e., <1.0, light 
distribution curves are obtained (Fig. 14) which indicate that the pro- 
jection screen is now darkest in the center. Here the crater images 
tangent inside the focal circle are too small to carry sufficient light to 
the center of the aperture. 

This particular type of reflector facilitates adaption of the optical 
system to any particular projector; for in contrast to the customarily 
employed elliptical reflectors, the angle of inclination represents an 
additional degree of freedom in design. 

2. In optical illumination systems for projection lamps one condi- 
tion is of paramount importance : the axes of all light beams passing 

Jlluminating Beam for P 


Figure 15. 

through every point, P, in the aperture must intersect at the center 
of the entrance pupil, c, of the projection lens (Fig. 15). 

In order to obtain a high efficiency of illumination, only such parts 
of the reflecting surface should appear luminous from a point, P, as are 
cut out of the reflecting surface by the extension of a cone having its 
apex at P, and with the pupil of the projection lens as its base. 

The new illumination system for the Ventarc lamps has been laid 
out to meet fully this requirement, in combination with a projection 
lens of 100-mm (4-in.) focal length. 

A very simple and conclusive test for this quality of the reflector 
may be made in the following way : the positive crater is replaced by 
a screen of homogeneous luminosity and photographs of the reflector 
are taken from points at the center and corners of the aperture, show- 
ing the portions of the mirror surface that appear luminous from these 
points. The theoretical form of the luminous areas on the mirror 




to be expected for each of these points may be calculated and com- 
pared with the actual form of the areas shown by the test photo- 
graphs. This method provides a convincing criterion for the quality 
of the reflectors. For instance, a measure of light distribution over 
the illuminated spot on the aperture plate is obtained directly from 
the ratio of the size of the luminous areas in the center to the size of 
such areas at the sides or corners. Figure 16 represents such a test 

Fig. 16. Test photograph. 

photograph showing the illuminated areas of a Ventarc reflector as 
seen from the center and corners of the film aperture. 

3. The requirement, D c -M a = D , can be met with reasonable ex- 
pense in the conventional coaxial optical system only for small diam- 
eters of the crater corresponding to arc currents up to 75 amp. In 
view of the minimum length of carbon which must be accommodated 
between the crater and the film aperture at higher currents, mirrors of 
very large diameter and excessive cost are demanded. 

To avoid such difficulties a deflected light path is employed in the 


Ventarc lamp with highest light output. Figure 17 shows the actual 
arrangement : the positive carbon is arranged vertically (in the lamp 
house) , the concentrating air blast is directed upward so that the nor- 
mal thermal rise of the flame gases adds to the effect of the blast. 
The optical axis of the illumination system is deflected by a mirror 
into alignment with the projection axis. The positive carbon passes 
through a slot in the deflection mirror, and its necessary length no 
longer affects the dimensions of the illumination system. 

As it is impossible to satisfy all desirable conditions such as homo- 
geneity of illumination on the aperture plate, optimum speed of the 
illumination beam and appropriate location of the plane of a dia- 
phragm by suitable formation of the reflecting surfaces alone, the 
foregoing system is designed to give an intermediate image (Fig. 17) 
much larger than the aperture. The subsequent introduction of an 

^^ Reflector 

f^ intermediate 

\T^ Crater Jmage Film Aperture 

v r"^^ 7 

Diaphragm Projection 

Plane Mirror Condenser |_ ens 

Positive Carbon 
Fig. 17. Optical system for Ventarc with ring-cathode. 

additional condenser lens is not objectionable, as this may be surface- 
treated to reduce reflection losses. The same is true of the deflection 
mirror which may be so constructed with a reflection index of 97%. 
4. The illumination system of a Ventarc lamp, designed for a crater 
diameter of 6.3 mm, produces a cone of light at the aperture of //1. 8 
speed. A similar system designed for a larger crater would only in- 
crease the speed of the light beam at the aperture, and this would be 
useless with present-day projection lenses, already filled by an //1. 8 
cone. A higher light output may be attained only by increasing the 
brilliance of the positive crater, i.e., by increasing the current density 
and consequently the consumption of the positive carbon. The lay- 
out of the Ventarc lamp with ring cathode provides for this possibility, 
as the essential dimensions of the interior parts and of the lamp 
house offer sufficient space to permit an increase of the positive carbon 
loading to 150 or 200 amp. 





1. A most important requirement for a good projected picture is 
effective luminous flux arriving at the projection screen. Assuming 
uniform illumination of the aperture, the following formula is valid 
for the luminous flux, L, illuminating the film : 

L = l-F-Q 

where I represents the brilliancy of radiation in the plane of the 
aperture measured in candles per mm 2 ; F, the area of the aperture 
in mm 2 ; and 0, the solid angle by which the pupil of the projection 
lens is seen from the film. An evaluation of this very simple relation 
is represented by Fig. 18, in which shadowing caused by the positive 
carbon guide mechanism is neglected. 

150 000 


too ooo 


Brilliancy [c/mm*] 



V \ \ 



\ \ 

\ \ 


F- 320mm 2 -\ 

Aperture 1=1 H,4 1=2 

Opening Angle of illuminating beam 

Fig. 18. Lumens through the film aperture. 

It is not, however, the peak value of brilliance measured at the 
most advantageous angle and at the brightest spot of the positive 
crater which is significant here, but rather the average of the brilliance 
over the crater surface at all angles. This average brilliance, which 
determines the luminous flux through the aperture, is a considerably 
smaller fraction of the peak brilliance than is generally presumed. 
For instance, the assumption of a homogeneous illumination of the 
image field with an average brilliancy of only 750 c/mm 2 , leads to the 
amazing value of 46,000 1m in the aperture plane. 

The figures of peak value of brilliance in the positive crater gen- 
erally contained in technical publications ought to be supplemented 
by a specification of the average brilliance all over the utilized surface 




of the crater, in order to avoid false impressions of the performance of 
different light sources. 

2. The heat capacity of the film is limited, and may not exceed a 
maximum value of 0.5 W/mm 2 even if forced air cooling of the film is 
employed. It is therefore necessary to restrict the luminous flux 
to the useful luminous region of the spectrum, so that maximum 
lumens through the aperture are obtained without exceeding the ad- 
missible radiation density. 

3. Next to uniform distribution and constancy in time, uniformity 
in color of the screen illumination is of great importance. Any varia- 
tion of color across the screen is decisively detrimental to the artistic 
value of projection. 

A very good criterion for the uniformity in color of the screen illu- 
mination is obtained by measuring the relative value of the green and 
the red component at the center and the corners of the screen by 
means of a phototube and color filters. 

Fig. 19. Red-to-green ratio of the screen illumination for the Ventarc, 50 amp. 

The results obtained with the Ventarc 50-amp are indicated in 
Fig. 19. It must be emphasized that with Ventarc lamps the dis- 
tribution of such values is rotationally symmetrical, whereas with 
customary high-intensity lamps symmetry only exists relative to a 
vertical line through the center of the picture. Extension of such 
measurements beyond the corners of the screen is recommended, 
as this permits prediction of the extent to which the red-and-green 
component in the corners will deviate if slight maladjustment of the 
projection system occurs. 

In view of the considerable importance of color uniformity, stand- 
ard tolerances should be set up for the red-to-green ratio on the 
projection screen. 


The light output of the Ventarc lamps is illustrated by Fig. 20. 
The types of the series hitherto developed cover an effective screen 




illumination range from 3550 up to 30,000 1m with arc currents of 20 
to 100 amp. All such lamps are equipped with reflector illumination 

The ratio of screen lumens to power consumption may be employed 
as a quality coefficient of a projection lamp (Fig. 20). An average 
coefficient of 6 Im/w is obtained by the Ventarc lamps where the 

Lm10 J 



20 W 60 80 tOO 120 \kQ 160 Amp 

Fig. 20. Screen lumens of the Ventarc lamps side-to-center ratio 80%. 


Water cooling 

Fig. 21. Positive head. 

brightness at the lateral boundaries of the screen amounts to at least 
80% of the value at the center. 


1. The positive head (Fig. 21) is water-cooled and the carbon sup- 
port and contacts for current supply are in good thermal contact with 
the water-cooled parts. As no decrease of the brilliance is observed 
after a new carbon has properly burned in, direct water cooling of the 




current contact is dispensed with as that would require considerable 
complication of design. As far as a reduced consumption rate can be 
achieved by using the special carbons of National Carbon, direct 
water-cooling of the jaws can be provided. 

The magnet coil is mounted behind the water-cooled parts of the 
positive head. It is thus protected against direct radiation and its 
current loading may consequently be made exceedingly high. 

The blower nozzle is made of a special acid- and heat-resistant 
steel. The nozzle may easily be replaced, but under conscientious 
operation the attainable period of life is very long. 

The air velocity may be regulated by means of a valve so that the 
optimal value for any current load of the positive carbon may be set 



Driving and 



Fig. 22. Negative ring system. 

To avoid excessive irradiation and heating of the lamp house by the 
highly intense anode flame, the water-cooled positive head is equipped 
with a cylindrical diaphragm, provided with the necessary ports for 
the observation of the crater and of the positive feed control. Heat- 
ing of the lamphouse is effectively reduced by this diaphragm. 

2. The ring cathode poses some constructional problems (Fig. 22), 
as, for instance, the mounting, drive and current supply of the cathode 

The two rollers previously mentioned, which guide the ring at sub- 
stantially constant arc length, are shown in the vicinity of the positive 

A symmetrical application of the current to the arc through both 
halves of the ring id essential as distortion of the arc would result from 




the magnetic fields excited by an imsymmetrical admission of current. 
Facilities permitting an easy replacement of the cathode ring are of 
practical importance; the rear wall of the lamphouse may be opened 
as a whole for the mounting of a new ring. 

3. The cathode ring and deflection mirror are arranged so that the 
ring completely embraces the mirror. Shadowing of the illuminating 
beam is thereby minimized. The somewhat higher shadowing occur- 
ring at the screen center improves the uniformity of screen illumina- 

4. Compressed air is provided by a blower separated from the 
lamphouse. The particular construction of the lamp (Fig. 23) permits 
utilization of one single blower to supply air alternately to two lamps. 


Lamp Housing 

Air Jeh for Pos. Head 
Fig. 23. Air jet system. 

This solution presents important technical and economical advan- 
tages, namely: 

(a) The hot flame gases do not pass through the blower so that 
difficult technical problems, e.g., excessive heat, choking by deposits, 
etc., do not arise. 

(b) The blower runs on cold air thus obtaining a higher efficiency. 

(c) The blower is mounted separately from the lamp, so that con- 
structional complications such as spring mounting, silencing, heat 
insulating and access for cleaning purposes are avoided. 

(d) Since only one blower system is required for two lamps, this 
may be built of higher quality. Silencer and dust filter are required 
only once, but simultaneously serve for two lamps. The diameter 
of the blower wheel may be increased and the rotation speed lowered 
as no restriction of the space exists. Consequently the commutator 




motor hitherto required for Ventarc lamps may be replaced by a 
squirrel-cage motor which is superior on account of its sturdiness and 
service-free operation. 

Because of the fixed position in space of the ring cathode, the 
Ventarc lamp provides additional safety against damaging of the 
positive head, in case of failure of the positive feed control. In such a 
case, the positive carbon would burn back until the arc was extin- 
guished by the minimum current relay or by the air blast. 


The particular form of the discharge of the blown arc permits deter- 
mination of the absorption rate of the arc plasma A,, the volumetric 
brilliance, &, and the saturation value of brilliance, b msLX , for increasing 



Fig. 24. Blown arc: volume brilliancy and absorption. 

diameters of the carbon assuming constant current density in the 
positive crater. This saturation value in the interesting range rises 
proportionally with the arc current. By measuring the values of 
brilliancy, 61, b z and 6 3 at the respective points, PI, Pz and P 3 in a 
direction inclined by an angle, 0, to the axis of the carbon, a very 
simple calculation arrives at the following relations (Fig. 24) : 


s ' log < 

6, = A 



61 - (62 - 6) 

- (6 2 - 6 3 ) 


S stands for the length of the line of sight through the arc stream 
looking towards points P 2 or P 3 . 

These relations hold for a very shallow crater as indicated by Fig. 
24. In case of a deeper crater the value bi has to be replaced by bi* 
% (&4 H~ W which is self explanatory with regard to Fig. 24. 

By inserting the particular values measured with a 200-amp blown 

b, = 700 6 2 = 1000 6 3 = 950 [IC/mm 2 ] 

the following figures are obtained : 

A = 0.25 

b v = 256 [IC/mm 3 ] 
6 ma * = 1025 [IC/mm 2 ] 

In this case the absorption of a layer of the arc plasma 1 mm in thick- 
ness amounts to 22%. The absorption rate of the arc plasma rapidly 
increases towards the region of short wavelength. This observation 
provides an explanation for the occurrence of brown areas inside 
the plasma at high current-density. 

NOTE: A 100-amp model of the new lamp was set up and demonstrated at the 
SMPTE Chicago Convention. 

The High-Speed Photography 
Of Underwater Explosions 



SUMMARY: A brief review of the techniques used in photographing under- 
water explosions at the Underwater Explosives Research Laboratory and the 
Naval Ordnance Laboratory for the past several years will be given. The 
Fastax (35-Mm), Eastman Hi-Speed and a rotating-mirror frame camera have 
obtained pictures ranging in speed from 2000 to 30,000 frames/sec. Ex- 
plosions of charges weighing up to one pound have been photographed at 
depths down to two miles. In the very deep water photography the equip- 
ment which synchronized the explosion with the flashbulbs, timers, etc., was 
self-started by means of a pressure switch. Typical results will be shown. 

MANY TYPES of transient phenomena have been studied by means 
of high-speed photography. In the past few years, under- 
water explosion phenomena have been included in such studies at a 
number of laboratories, notably the David Taylor Model Basin in 
this country and the Naval Construction Research Establishment in 
Great Britain. 1 Presented here is a brief outline of the work by a 
number of people at the Underwater Explosives Research Laboratory 
(UERL) and the Naval Ordnance Laboratory (NOL) where methods 
have been developed for the photography of explosive charges rang- 
ing in size from 1 oz to 300 Ib. The techniques used by other labora- 
tories, in most cases, have restricted the explosion to that which can 
be contained in a tank and consequently the charge size has usually 
been about 1 gram. 

The work described here has all been conducted in the ocean from a 
ship. In general, two types of high-speed photography have been 
used in the UERL studies of underwater explosion phenomena: 
(1) the photography of shock waves 2 by means of a short-duration 
light source (10~ 6 sec); and (2) the photography of explosion bubble 
expansion and contraction by means of motion pictures. 3 

As is well known, an underwater explosion results in a rapidly ex- 
panding shock wave followed by a succession of pressure pulses. 4 
The later pressure pulses are caused by the repeated collapse of the 
gas globe formed by the hot expanded gaseous products of detonation. 

PRESENTED: April 26, 1950, at the SMPTE Convention in Chicago. This paper 
is based in part on work done for the Bureau of Ordnance under Contract NOrd 
9500 with the Woods Hole Oceanographic Institution, Woods Hole, Mass., and in 
part on work at the Naval Ordnance Laboratory. 



It is the photographing of the oscillations of this gas bubble with 
which the following is concerned. 

Some of the requirements of such photography are: (1) sufficiently 
clear water to permit the proper spacing of equipment and explosive 
charges; (2) appropriate cameras contained in water and explosion 
resistant cases; (3) synchronization of the lights and camera with the 
detonation ; (4) light sources ; (5) precise speed control of the camera 
or timing marks on the film; and (6) a rig for mounting and main- 
taining the position of each component. 

The requirement of very clear water results from the camera-to- 
object distance required in some of this work. For example, using a 
2^-in. focal length lens on the Eastman Hi-Speed camera, the angle 
of view under water, is only about 5 deg and the target must be 40 ft 
away to get a 4-ft field. Sufficiently clear water was found off the 
Florida coast near the Bahamas and in the Caribbean Sea. In 
general, clear water can be found in the tropical latitudes where it is 
not contaminated by shore drainage. A convenient measure of the 
water clarity is obtained by means of a Secchi disk which is simply an 
8-in. white circular disk. The limiting depth to which it can be 
lowered in the sea and remain visible is a reliable and reproducible 
measure of clarity. 5 - 6 The waters mentioned had a Secchi disk read- 
ing ranging from 135 ft to 160 ft. A rough rule for such photography 
is that adequate resolution can be obtained for a camera-to-target 
distance equal to about one-half the Secchi disk reading. 

The cameras used in this work included the Eastman Hi-Speed, 
the 35-Mm Fastax and a rotating-mirror frame camera of NOL 
design. These cameras were shock-mounted in a heavy, watertight 
case to prevent damage from an explosion. Figure 1 shows the case 
used with the Eastman camera. The NOL designed camera, which 
was used at depths as great as 2 miles in the ocean, was enclosed in a 
spherical case which was 22 in. ID and had a lM-in. wall. The 
camera lens viewed these deep explosions through a 1-in. thick window 
covering a 1/4-in. hole. 

The camera designed by S. J. Jacobs (NOL) and A. A. Klebba 
(Woods Hole Oceanographic Institution) is essentially a modified 
Bowen camera. The image is focused on a spinning mirror which 
has the focal axis of the taking lens system for its axis of rotation. 
The plane of reflection of the mirror is 45 deg to this axis. The 
image is thus reflected through the framing lens to the stationary film. 
One hundred framing lenses provide 100 pictures. With the mirror 
revolving at the rate of 18,000 rpm, 100 pictures can be taken at the 
rate of 30,000 fps. An auxiliary shutter prevented multiple exposure. 





Fig. 1. High-speed camera and "explosion proof" case. 


Such frame speeds were required for the very deep photography in 
which the oscillations of the explosion bubble are much more rapid 
than in shallow water. The Eastman and Fastax cameras were not 
used at depths greater than 1,000 ft. 

The light source most commonly used consisted of a number of focal 
plane flashbulbs, having a duration of about 70 msec (milliseconds). 
On occasions when a longer light was necessary, such flashbulbs were 
fired in series. Miniature bulbs were used for the very deep work. 
It was necessary to provide protection from the explosion for these 
bulbs by a watertight case having a Lucite window. 

For the work at less than 1,000 ft. with the Eastman and Fastax 
cameras electrical power to operate the cameras and a timing signal 
was provided through cables from the ship. Since knowledge of the 
precise depth at which the experiment was performed was very im- 
portant, a depth gage was used to measure the static pressure and 
hence the depth to within 2%. For some experiments a Bourdon 
gage was photographed by a small camera at the instant of the ex- 
plosion. For others the Bourdon gage element actuated a poten- 
tiometer which was one arm of a Wheat stone bridge. The latter 
method permitted continuous reading by means of cables to the sur- 
face and consequently was more convenient for an accurate pre- 
setting of the depth. 

It was impossible to have electrical cables leading from the camera 
to the surface for work at depths of one and two miles. The power 
for the camera in this case was supplied by miniature wet cells, and 
the entire equipment was self-operated after the closing of a depth 
switch. The sequence of events necessary to obtain photographs 
included : the starting and stopping of the camera motor, the opera- 
tion of the shutter, the firing of the 1-lb explosive charges, and the 
firing of 20 to 40 No. 6 focal plane flashbulbs. These events, which 
had to be synchronized with a precision of the order of M msec, were 
started by the closing of a switch triggered by hydrostatic pressure. 
Two pressure switches, one actuated by a Bourdon gage mounted in 
the camera case and one which involved the movement of a spring- 
backed piston, had an accuracy of better than t =*=H% i n depth. 

Since the Fastax camera did not have a built-in firing switch, it 
was necessary to fire the charge from the laboratory ship. In order 
to fire the charge automatically when the camera had reached maxi- 
mum speed and the photoflash bulbs were near peak intensity, and to 
turn off the camera, the firing circuit and camera power circuit were 
combined as shown in Fig. 2. 

A typical operation would be as follows : when the safety switch is 







thrown and the firing button on the Time-0-Lite* is pressed, a 110-v 
current is transmitted to the camera power relay and trips the elec- 
tronic time delay. 3 The camera then starts and is up to speed in a 
little more than a second. At 1.5 sec, the time delay gain changer 
closes an internal relay which transmits the 110-v a-c power to the 
firing relay on the rig. This fires the flashbulbs and the charge, the 
latter being delayed a few milliseconds by a series resistor so that the 
bulbs can be at maximum brilliance when detonation occurs. After 
3 sec, the Time-O-Lite turns off the power to the camera relay, thus 
stopping the camera after the film has run through. These times may 
easily be changed to suit other conditions. 

As shown in Fig. 3, all the necessary components in their respective 
pressure cases were mounted on a rigid frame. Such steel frames, 






3' 6" SQUARE 

Fig. 3. Deep sea camera apparatus. Electrical connections not shown. 
Dimensions: 19 ft 6 in. long, over-all; 3 ft 6 in. square. 

built of channel and angle iron in lengths varying from 10 ft to 40 ft, 
were used for explosive charges weighing up to about one pound. 
Since the operation of the camera was automatic with the closing of 
the depth switch, this entire rig, weighing about 1,200 Ib, was simply 
lowered to the preset firing depth by a single steel cable. 

In Fig. 4 are shown some typical results using the Eastman Hi- 
Speed camera. Here is shown the explosion of 25 grams of tetryl at 
a depth of 350 ft. The scale above the charge is 12 in. long. The 
charge was placed 20 ft from the camera where the field of view was 
about 20 in. X 28 in. The first frame shows the undetonated cylin- 
drical charge. Detonation occurred between the first and second 
frames. The bubble, in this case, grows to its maximum size in 5.5 
msec and has collapsed to a minimum at the end of 11 msec. The 

* Time-O-Lite Master Model M-49, Industrial Timer Corp., Newark, N.J. 




Fig. 4. Explosion of 25 grams tetryl at 350 ft. 

successive expansions and contractions are clearly shown. The un- 
symmetrical shape of the bubble at its minimum is typical of such ex- 
plosions in shallow water and is the result of carbon particles left 
behind in the water by the collapse of the gas globe. Front-lighting 
was provided by four No. 31 GE photoflash bulbs placed midway 
between the camera and the charge. A white background was used 
to silhouette the object. The exposure was made at f/2.7 and the 
camera speed was 2200 fps. The film used was Super-XX Panchro- 
matic Negative. Twice the normal development time in D-76 pro- 
duced reasonable contrast. The camera was focused by means of 
a ground glass in air, taking a distance in air three-fourths of the 
object distance under water. 

An oscillating bubble such as is produced by an underwater ex- 
plosion has interesting characteristics when it is near a surface. Such 
a bubble is repulsed by a free surface and is attracted by a rigid 
surface. For example, the bubble from a 1-oz charge when fired 
just beneath the water surface will go down instead of up. An ex- 
ample of a shallow water explosion is shown in Fig. 5. This photo- 




Fig. 5. Explosion of 25 grams tetryl charge 1 ft 9 in. beneath surface. 

graph was obtained by using the 35-Mm Fastax camera. Here back- 
lighting was used and two sets of No. 31 photoflash bulbs were set off 
in sequence in back of a translucent screen behind the charge. The 
circuit used to fire the flashbulbs in sequence, 50 to 90 msec apart, 




is shown in Fig. 6. The 25-gram tetryl charge was 1.75 ft beneath 
the surface. Small cavitation bubbles resulting from the passage of 
the shock wave partially obscure the early frames. At its maximum 
size the bubble is approximately tangent to the water surfaces. As 
the bubble collapses there is pronounced interaction with the surface. 
A large tubular connection to the atmosphere permits air to enter the 
bubble in which the pressure is less than atmospheric when it is close 
to its maximum. Enough air thus enters the bubble so that its 
volume at minimum size is larger than for the deeper shots. The 
tubular space between the bubble and the surface is apparently not 
a continuous open tube after the minimum, since the bubble once 
more expands and does not vent. 

H|i|ih oR " 



Fig. 6. Sequence firing circuit for two sets of flashbulbs; 
Underwater Explosives Research Laboratory, Woods Hole, Mass. 

An example of the results using the NOL camera to photograph 
very deep explosions is shown in Fig. 7. This shows the explosion 
of a 1-lb spherical charge at a depth greater than two miles. Again 
the first frame shows the unexploded charge which was 3Ji in. in 
diameter. The detonation light can be seen in the second frame. A 
linear scale is provided by the black squares in the lower part of each 
frame. The distance between the inside edges of the squares is 
12 in. and between the outside edges it is 18 in. The maximum and 
minimum bubble diameters are 12^ in. and 5^ in. respectively. 
The most notable difference in such deep explosions, as contrasted 
with shallower ones, is that the bubble more nearly retains spherical 




^^m k ^^M* ' m ;: V 

k 1 ik m>^., J L^. ..... .^ 

Fig. 7. Explosion of one-pound charge at 12,000 ft. 

symmetry throughout its cycle. The carbon particles from the ex- 
plosion have apparently not progressed outside the gas globe. This 
has permitted the accurate measurement of the variation of bubble 
volume with time which is of considerable theoretical importance in 
the study of explosion phenomena. The reproducibility obtainable 
from such photographs is illustrated in Fig. 8 in which is plotted the 
variation of bubble diameter with time for a 1-lb charge at a depth 
of 6,000 ft. 



SHOT NO. "155-12 


Fig. 8. Variation of bubble diameter with time. 

In conclusion it may be pointed out that high-speed photography 
of explosion phenomena has advanced in the last decade from a 
curiosity to one of the most valuable tools available for the study of 


1. D. A. Senior, "High-speed photography of underwater explosions," Phot. 

Jour., Section B, 86B, 25, 1946. 

2. J. E. Eldridge, Paul M. Fye and R. W. Spitzer, "Photography of under- 

water explosions, Pt. I," OSRD Report No. 6246, Mar! 1947. 

3. E. Swift, Jr., Paul M. Fye, J. C. Decius and R. S. Price, "Photography of 

underwater explosions, Pt. II, High-speed photographs of bubble phe- 
nomena," NavOrd Report No. 95-46, Dec. 1946. 

4. Robert H. Cole, Underwater Explosions, Princeton University Press, 1948. 

5. M. Ewing, A. Vine and J. L. Worzel, "Photography of the ocean bottom," 

Jour. Opt. Soc. Amer., vol. 36, pp. 307-321, June 1946. 

6. F. A. Jenkens and I. S. Bowen, "Transparency of sea water," Jour. Opt. 

Soc. Amer., vol. 36, pp. 617-623, Nov. 1946. 

A Heavy-Duty 
16-Mm Sound Projector 



SUMMARY: An intermittent sprocket pulldown with accelerated geneva 
drive has its own directly connected synchronous motor. The remaining 
sprockets and shutter are driven by a second synchronous motor. The two 
systems, engaged temporarily for starting, run mechanically independent to 
eliminate shock forces and obtain an inherently flutter-free sound drive. 
New optics throughout give high picture and sound resolution. Tungsten and 
arc light sources are provided. Other features include a high-quality ampli- 
fier, independently driven accessories, turret accommodation for instan- 
taneous lamp replacement, improved base-up mounting of lamp and im- 
proved floor mounting. 

old. From the start, it has found increasing application in 
organizational functions such as teaching and training, although it 
was inaugurated to provide personal motion pictures. During the 
last war, it was used extensively in training and recreational pro- 
grams, especially by the armed forces. Under hard military usage, 
projectors wore out rapidly and were generally unsatisfactory for 
that type of service. This led to the development of Joint Army and 
Navy specifications, JAN-P-49, for a projector that would be ade- 
quate for the services. A great deal of effort was put into these 
specifications, which were scaled to the ideal objective rather than to 
one that could easily be attained. A postwar review of the poten- 
tialities of 16-mm equipment soon disclosed that amateur projectors 
were not adequate for testing purposes and that little could be done 
in evaluating other 16-mm equipment and processes until a new pro- 
jector was available. 

The projector described in this paper was developed to meet the 
essential features of the JAN specifications. We believe that the 
intent of all the items is met and that in some respects the require- 
ments of the specifications are exceeded. This projector, named the 
Eastman 16-Mm Projector, Model 25, was designed to meet the de- 
mands of the growing civilian market as well as the needs of the 
military forces. Also, it was designed in anticipation of marked im- 
provements in the whole 16-mm process. 

PRESENTED: April 24, 1950, at the SMPTE Convention in Chicago. 







Fig. 1. Film threading path. 

. Electrical Linkage. A study of the problems of wear, noise, and 
flutter suggested that much might be gained in designing the mechani- 
cal system primarily to control the occurrence and magnitude of ex- 
traneous forces that give rise to these three troubles. Movement of 
the film intermittently in the gate requires a mechanism that has in- 
herently high accelerations. The forces generated are proportional 
to the masses that are accelerated, and their effect is felt in all the 
mechanism connected to the intermittent. The intermittent dis- 
turbances can be kept out of the other sections of the mechanism by 
filtering, just as flutter is reduced in conventional projectors. 

While the projector is running, there is no other acceleration. If, 
then, the intermittent is isolated and driven separately, the re- 
mainder of the mechanism is protected from these disturbances. 


Also, the system, including the shutter and the sprockets, can be de- 
signed for a minimum of flutter, flicker, noise and wear. The Model 
25 is so divided, each unit having its own motor. The sprocket- 
shutter mechanism, contained by the main casting, is driven by a 
four-pole 1800-rpm synchronous motor, which has its rotor mounted 
on the top end of a vertical shaft. For the intermittent mechanism, 
a special synchronous motor having a speed of 1440-rpm is provided. 

Since the intermittent is separated from the shutter, some means 
of exact phasing is necessary. To provide this phasing, two special 
synchro-gears mesh in the usual way during starting, but as the 
motors pull into synchronism, their teeth float clear of each other. 
This clearance is provided by removal of alternate teeth from con- 
ventional gears. To assure proper contact of the teeth, each of these 
gears is made of two such cutaway gears assembled side by side with 
the teeth of one opposite the spaces of the other. 

This principle of isolation of mechanical systems by electrical link- 
age is carried further by the use of three additional motors to drive 
the take-up spindle, the rewind spindle, and the blower. Each 
motor is chosen specifically for the particular job it has to do. 

Figure 1 shows the positions of the sprockets, gate and shutter 
housing. Figure 2 reveals the intermittent system complete with its 
motor and synchro-gear. In Fig. 3 we see this system assembled to 
the sprocket-shutter mechanism. This rear view shows the motor 
for the intermittent and its synchro-gear in mesh with an identical 
gear on the sprocket-shutter mechanism. In order to show these 
parts, the flywheel for the sound system was removed. 

Pulldown System. The pulldown mechanism comprises an inter- 
mittent sprocket assembly housed in a separate casting and driven 
directly by the specially designed 1440-rpm synchronous motor. 

Various means of driving the intermittent sprocket, including a 
"drunken" screw, were considered. However, the geneva star was 
found to be the most desirable. Here the star does not conform 
strictly with conventional motion picture practice, because it has 
eight positions instead of the usual four. A four-frame sprocket 
would be far too small for 16-mm work, and an eight-frame size 
seemed the best compromise with respect to the mass of the sprocket 
and the acceleration in the drive. Basically the driving angle of an 
eight-position geneva movement is 135 deg, but in this projector, it is 
accelerated to 57 deg by an off-center driving system. In order to keep 
bearing loads within reason, we divided this acceleration between two 
off center driving elements arranged in series, as shown in the skele- 







Fig. 4. Skeleton of intermittent system. 

ton system (Fig. 4). Viewing from left to right, we see the sprocket, 
the geneva and its driver, the two balanced accelerators, the synchro- 
gear and the motor. Inside the hub of the synchro-gear, there 
is a coiled-spring flexible coupling that has an important function. 
With this system substituted for a more rigid coupling, there is a ma- 
terial improvement in the quietness of the action. It would be ex- 
pected that such a flexing element would increase the time required 
for the pulldown action. However, the oscillating system formed by 
the spring coupling and the moment of inertia of its load may be 
tuned to produce the normal 57-deg pulldown. In a model of the 
projector adapted to television, this tuning is adjusted to produce an 
action somewhat less than 50 deg. 

The intermittent assembly is fitted with two mounting trunnions 
that are concentric with the sprocket. Thus the unit can be rotated 
for framing by means of the lever shown in Fig. 2. This lever en- 
gages an eccentric on a shaft that extends to an external knob. 

Because of the precipitous change from positive to negative 
acceleration, which is characteristic of the geneva movement, any 
gears with their necessary backlash in the drive train would be sub- 
jected to shock forces. Furthermore, the oscillatory nature of the 
drive through the flexible coupling would aggravate this trouble, but 
the use of the 1440-rpm motor eliminates the gearing that would 
otherwise be needed, thus precluding its troublesome backlash. 
Two of the three remaining elements that could introduce backlash 
are the slots and splines of the accelerators. Since they have an 




appreciable area of contact, which is not possible with gears, the film 
of oil between them effectively cushions these forces. Thus a quiet, 
durable mechanism is obtained. The third element is the pin of the 
geneva movement. This is made with high precision and has also 
the benefit of the oil bath. 

The 1440-Rpm Motor. The rotor of this motor has five poles. The 
stator has four poles, each wound in one-fifth of the circumference. 
At a given instant, the motor is polarized as shown on the right of 
Fig. 5. In a conventional motor, such as the one at the left in Fig. 5, 
once synchronous speed is achieved, the poles of the rotor maintain 
the same polarity unless the motor is overloaded. In this new 
motor, also, each pole of the rotor maintains the same polarity as it 
passes the four poles of the stator. But as it passes the open position 
in the stator, it "loses step" and is forced into a reversal of polarity 
as it faces the next stator pole. The time required to pass the four 
poles is ^20 sec > the same -as required for one revolution of a con- 
ventional 1800-rpm motor. However, the rotor has turned but ^ of 
a revolution during this time. Thus its speed is 1440 rpm. 

Sprocket-Shutter System. Ordinarily, the hunting characteristics of 
synchronous motors cause excessively large shock forces in gearing 
when parts, such as a shutter, having large moments of inertia are in- 
cluded in the mechanism. This condition, much as in the case of 
the intermittent, would cause noise, rapid wear of the gears, and 
flutter in the sound system. In this projector, a protection is 

1800 R. P.M. 

Fig. 5. Diagram to show 1440-rpm motor action. 


afforded by a spring coupling between the shutter and its shaft. 
Thus the reaction torque is reduced, and the hunting of the motor is 
suppressed. The single-bladed shutter runs at 2880 rpm to give 
two interruptions of the light per frame of film. It was placed as 
near the gate as possible in order to provide the minimum angle for 
covering the light beam. 

All three continuous sprockets feed twelve frames of film per 
revolution. The film is guided with minimum side clearance through 
these sprockets. They are provided with separate, spring-actuated 
rear flanges that permit free passage of splices that are not aligned 
perfectly laterally. All sprockets are of hardened steel. 

The problem of meeting all requirements in constructing the gate 
is perhaps the most difficult in designing any projector. It always 
seems necessary to make some compromises. We use projection 
lenses that will resolve at least 90 lines/mm over the whole frame if 
the film is flat. These lenses were described by W. E. Schade 1 at 
the Fall 1949 meeting of SMPE. However it is a problem of long 
standing to hold -a film flat against its normal curl and against the 
thermal distortions caused by the light beam. Here a compromise 
has been made in curving the gate to a radius of 3 in. When the 
focus is adjusted to obtain the best average definition for the center 
and the edges of the screen, a dynamic resolution of 60 lines/mm 
is obtained. This curved construction also insures against scratching 
of the film, caused by contact between the gate and the picture and 
sound areas of the film. For threading, the gate is moved forward 
with the lens, and it returns to position with the focus of the lens 
undisturbed. As in framing, focusing is controlled by a knob on the 
outside of the case. The pivots of this linkage are spring-loaded 
to take out backlash, and the knob is provided with a lock. Four 
sapphire pads are used for side-guiding the film. 

General Structure. The lamphouse is mounted on a casting that 
encloses the blower, which is mounted directly on the shaft of a 1725- 
rpm induction motor. This assembly is mounted on a base casting, 
as are the assemblies containing the intermittent motor, the main 
mechanism, and the pre-amplifier. An enclosing case is carried by 
the same casting. The supply arm mounted on the case completes 
the projection-head assembly. This whole combination is sup- 
ported on a pedestal-cabinet that places the projector the correct 
distance from the floor. This is shown in Fig. 6, which shows the 
projector adapted for arc illumination. If the arc lamphouse and 
the rectifier are covered, the appearance is that of the tungsten model. 

A platform pivoted about a hinge pin at the top rear edge of the 




Fig. 6. Complete arc model of projector. 

pedestal holds the projection head. It can be tilted, by means of a 
jackscrew inside the front of the pedestal, for projection to a screen 
located above or below a horizontal plane through the lens. Two 
positions for the hinge pin provide two ranges of tilt. The pedestal 
also supports the take-up arm. 

Tungsten Illuminating System. A base-up, 1000-w, 10-hr lamp with 
an improved basing ring is used. This new method provides more 
accurate alignment of the filaments, better candle-power main- 
tenance, and improved ventilation. A dual lamp support, Fig. 7, 
permits rapid replacement of a burned-out lamp. Without stopping 
the projector, the operator can quickly swing out the old lamp and 
move a new one into place by means of the lever shown. Then the 




Fig. 7. Turret holder for tungsten lamps. 

defective lamp can easily be removed from the top for replacement. 
A specially designed condenser provides the //1. 5 cone of light needed 
for the objective lenses. 

Arc Illuminating System. For arc illumination, the tilting plat- 
form is replaced by one that is extended to carry the lamphouse for 
the arc. Of special design and styling, this arc lamp is made by the 
Strong Electric Co. A condenser that supplements the mirror is 
placed in the same position as the condenser for the tungsten lamp. 
A heat filter located immediately in front of the lamphouse is essen- 
tial for black-and-white film, but it can be swung out of position for 
color film. 

434 EDWIN C. FEITTS October 

The Sound System. Three rather distinct functions characterize a 
sound system. First, the film must be moved with a high degree of 
uniformity of motion and in a precise location past a scanning posi- 
tion. Second, the film must be scanned with a light beam of proper 
dimensions. Third, the modulated light from the film must be con- 
verted into a modulated electrical signal, and this signal must be 
amplified for reproduction. 

From the foregoing discussion, it will be seen that particular pains 
have been taken to meet the first requirement, in that the film- 
driving mechanism was designed as an inherently flutter-free system. 
It is isolated completely from disturbances originating in the inter- 
mittent movement, and the spring coupling for the shutter minimizes 
the effect of hunting in the motor. Other features provided to meet 
the requirement include the following: the linkage between the 
motor and the sound sprocket is through a single-stage worm and 
worm wheel; the ball bearings carrying the sound drum shaft are 
mounted in a quill to assure optimum alignment; and the sound 
sprocket is designed after Chandler 2 to minimize flutter arising from 
a lack of match in pitch between the sprocket and film. Thus 
the system is designed throughout for a minimum of flutter. 

Between the sound drum and the sound sprocket the film passes 
over an idler mounted on a spring-held arm. The oscillation of this 
arm is viscously damped by a film of silicon fluid. Tension in the 
loop is provided by an eddy-current drag between an aluminum disc 
on the flywheel and a fixed permanent magnet. A separate take-up 
sprocket provides a free loop on the other side of the sound sprocket 
in order to keep the functioning of the sprocket constant as it feeds 
the film from the sound loop, and to protect the sound loop from dis- 
turbances originating in the take-up. 

The "slitless"-type sound reproducer is essentially as described by 
McLeod and Altman. 3 An image of the filament of the exciter lamp 
is formed on the film. An intermediate image formed by a special 
curved cylindrical element is free of filament character. It is 0.05 
of the width of the source, is curved toward the objective, and is 
limited in length by a suitable field stop. This intermediate image, 
in turn, is focused on the film, at a further reduction of 3 to 1 by a 
highly corrected microscope objective, as a flat image that is uniform 
in width and in light intensity throughout its length. 

The last point is necessary in order to reproduce variable-area 
tracks at reasonably low distortion values. Also, the level of illumi- 
nation must be sufficient to give a high signal-to-noise ratio. Our 
measurements of distortion are in terms of intermodulation obtained 




with especially prepared, variable-area test films. Since it is diffi- 
cult to evaluate the width of the slit by straight optical methods, the 
equivalent value is derived from the amount of attenuation of the 
high frequencies on SMPTE 16-Mm Multifrequency Test Film. 
The reference level for noise measurements is obtained from the 
SMPTE 400-Cycle Signal-Level Test Film. 

It is well known that the reproduction characteristics of com- 
mercially produced 16-mm films vary over a wide range. In the 
absence of precision projection equipment, it has been difficult to 
demonstrate the ultimate quality of either picture or sound. This 
projector is designed for best reproduction of films made according to 


Fig. 8. Response versus frequency curves. 

the most advanced production practice. At the same time, it is 
necessary to handle the full range of quality encountered in films 
coming from current production. Added to this is the problem of 
the great variations in listening conditions encountered in 16-mm 
operation. A wide range of equalization to meet these conditions is 
provided by step-switch controls on both the low and high fre- 
quencies. The curves for these steps shown in Fig. 8 represent the 
response from SMPTE 16-Mm Multifrequency Test Film. 

Electronic System. The electronic system is divided between a pre- 
amplifier on the projector head and an assembly mounted in the 
pedestal, with controls as shown in Fig. 6. In the latter assembly, 




Fig. 9. Amplifier and power supply system. 

there is the main amplifier and a d-c power supply for the exciter 
lamp and the heaters in the pre-amplifier. Figure 9 shows the com- 
plete electronic system dismounted and open for servicing. This 
equipment was engineered in cooperation with the Altec Lansing 
Corp. and is manufactured by them. It is designed primarily to drive 
their 604B or 800 loudspeaker. The 604B is supplied in a special 


cabinet styled to match the projector, but the 800 is their standard, 
unmodified speaker. Altec Service Corp. will supervise the installa- 
tion of the equipment and offer their facilities for servicing it. 

Controls. The main power switch connects the system in the follow- 
ing sequence : 

1 . All circuits are off, but it is possible to energize the rewind cir- 
cuit by means of the lever on the take-up arm. 

2. Blower, rewind, and take-up are on, the last two at reduced 
voltage, as described below. 

3. Blower, take-up at full voltage, and the two motors for the 
mechanism are on. 

4. To point 3 is added the projection lamp for full operation. 
After the projector is threaded and the switch is set on position 2, 

the torque motors located in the two arms are operated at reduced 
voltage to take up all slack film and pull it gently against the sprockets. 
This part of the threading, therefore, is assured before projection is 
started. At the take-up, this elimination of the slack minimizes the 
possibility of damaging the film as it tightens. At the supply spindle 
it serves the same purpose, and it also prevents overrunning of the 
supply reel. Both the torque and the speed of the take-up are ad- 
justed to accommodate a minimum reel-hub diameter of 4^ in. 
Operation of the lever on the take-up arm disengages the take-up 
motor and puts full voltage on the rewind motor in the upper arm. 

Lubrication. Lubrication of the intermittent is by an enclosed 
splash bath, while the shutter-sprocket mechanism is flooded with 
oil from an impeller pump on the bottom of the vertical shaft. Ball 
bearings are used whenever feasible, and all other bearings are de- 
signed for infrequent attention. 


Lenses: Focal lengths of 2, 2%, 2%, and 3 in. All luminized and all //1. 5. 
Resolution of lens: At least 90 lines /mm over a flat field. 
Dynamic Resolution: 60 lines /mm. 
Light Source: Tungsten 1000-w, 10-hr lamp, 
Improved base-up mounting, 

Arc, 46-amp, nonrotating, positive, high-intensity, with 7-mm posi- 
tive, and 6-mm negative copper-coated carbons, 
11-in. mirror with supplementary condenser lens. Carbon trim for 

80 min. 

Shutter transmission: 60% with two interruptions per frame. 
Illumination: Shutter running, no film; 
Tungsten: 4501m; 

Arc: With heat glass for black-and-white film, 20001m; 
Without heat glass for color film, 250Q 1m- 

Steadiness: Horizontal unsteadiness less than 0.00025 in.,. 
Vertical unsteadiness less than 0,0005 in. 



Flutter: Less than 0.2% nns. 

Equivalent Slit Width: 0.0003 in. 

Inter modulation Distortion: 5% including scanning beam and amplifiers. 

Signal-to- Noise Ratio: 55 db, using SMPTE 400-Cycle Signal-Level Test Film. 


Elevation: 10 deg above horizontal, 17 deg below horizontal. 
Weights: Projection head 90 Ib, 

Pedestal with amplifier 165 Ib, 

Complete tungsten model 255 Ib, 

Complete arc model with rectifier 640 Ib. 
Dual Projection: Common amplifier and instantaneous change-over device. 


The ultimate technical possibilities of 16-mm cinematography are 
far from being realized. But as each step of forward progress is 
made, we can see that the potential uses of more highly developed 
equipment and processes are great indeed. By careful selection of 
films this projector will demonstrate the quality attainable in in- 
dividual items such as picture definition, steadiness and sound resolu- 
tion. At present it is seldom that one film has high quality in all 
these respects, largely because of the lack of adequate equipment at 
some point in the production of the film. However, much work is 
now being done, and the day may not be far off when 16-mm motion 
pictures come into their own for more serious semiprofessional and 
professional uses. We believe that this projector, in all functional 
respects, now raises the limits imposed by projection apparatus above 
those set by other equipment and processes. We offer it as a pre- 
cise tool to be used in bringing to pass this day of realization for 16- 
mm motion pictures. 


1. W. E. Schade, "A new //!. 5 lens for professional 16-mm projectors," Jour. 

SMPTE, vol. 54, pp. 337-344, Mar. 1950. 

2. J. S. Chandler, "Some theoretical considerations in the design of sprockets for 

continuous film movement," Jour. SMPE, vol. 37, pp. 164-176, Aug. 1941. 

3. J. H. McLeod and F. E. Altman, "An optical system for the reproduction of 

sound from 35-mm film," Jour. SMPE, vol. 31, pp. 36-45, July 1938. 

Interference Mirrors 
For Arc Projectors 



SUMMARY: A large fraction of the radiation from an arc lamp consists of 
infrared and ultraviolet energy. By coating the arc mirror blank with multi- 
layer interference films instead of with silver, the major portion of the visible 
light can be reflected while most of the infrared and near ultraviolet radia- 
tion is transmitted. Hence, this type of mirror reduces film distortion 
caused by overheating, and because of the selective nature of the reflection, 
it affords a means of controlling the color quality of the illumination. 

ONE OF THE BASIC PROBLEMS that must always be faced in the 
design of high-intensity projectors is the possibility of over- 
heating and buckling the film. This difficulty is particularly acute in 
the case of motion picture film projected by arc lamp illumination, 
where a great deal of energy is concentrated on a small area of film. 
The customary approach to this problem is to absorb the troublesome 
infrared radiation with a heat-absorbing glass placed between the 
light source and the film. 

Another approach to this problem is the use of multilayer inter- 
ference filters to separate the visible light from the infrared and ultra- 
violet radiation. Interference filters designed to transmit visible 
light and reflect infrared and ultraviolet have proved useful, but their 
effectiveness is restricted by the limited band of infrared energy 
which they reflect. Hence, such filters are usually less efficient than 
heat-absorbing glasses. They are useful, however, as supplementary 
filters or in applications where high efficiency is not necessary and 
where breakage of heat-glass is a major problem. 

Much more nearly complete separation of light and heat can be 
obtained if a multilayer filter is used in the inverse manner, that is, to 
reflect the light and transmit the infrared and ultraviolet radiation. 
The mirror in an arc lamphouse is particularly well suited for this 
purpose. If the mirror blank is coated with interference films instead 
of with silver, most of the visible light can be reflected, and most of 
the infrared and near ultraviolet radiation transmitted as illustrated 
in Fig. 1. 

PRESENTED: April 24, 1950, at the SMPTE Convention in Chicago. 


440 G. J. KOCH October 

The interference films shown in the diagram consist of four to eleven 
layers of transparent or semitransparent materials of alternately high 
and low refractive index. There is a wide variety of materials, in- 
cluding dielectrics and semimetals, that have optical properties suit- 
able for this purpose. Magnesium fluoride or sodium aluminum 
fluoride are the customary low-index materials, and zinc sulfide or 
titanium dioxide, the usual high-index materials. The thickness of 
the films is controlled so that visible light is reflected from each film in 
phase with the light reflected from the other films; hence, the com- 
bination produces high reflection. In general, however, for wave- 
lengths outside the visible region, the component reflections are out of 






Fig. 1. Simplified diagram showing position of arc, mirror and gate. 

phase, which results in low reflection and high transmission. Thus, 
the multilayer coating is a selective mirror that reflects visible light 
and transmits infrared and near ultraviolet radiation. 

Figure 2 includes a curve showing the approximate reflectivity 
versus wavelength for an interference mirror of this type. A second 
curve showing approximately the spectral reflection of a silvered 
mirror is shown for comparison. It reveals the large reduction in 
reflection in the infrared and near ultraviolet parts of the spectrum 
secured with the interference mirror. In fact, for infrared and near 
ultraviolet radiation, the thin films act as low-reflection coatings, 
allowing those regions of the spectrum to pass through the mirror 
with little attenuation. But for visible light the films act as high- 
reflection coatings with a reflectivity equal to or greater than that of a 
silvered mirror. 




The exact shape of the reflection-wavelength curve depends pri- 
marily on the following factors: the reflectivity at each film inter- 
face, the number of layers, the optical thickness of each layer relative 
to that of the other layers, the dispersion and absorption characteris- 
tics of the materials, and the angle of incidence of the light. The 
calculations involved in the design of efficient combinations are so 
formidable that graphical and analogue computing devices have 
proved most helpful. 

The films are deposited on the glass mirror blank in a high-vacuum 
coating system. The procedure consists in successively evaporating 




700 1000 2000 



Fig. 2. The spectral reflectance of a multilayer 
interference film compared with that of a silver film. 

the required number of layers on the glass blank, which rotates con- 
tinuously in the high-vacuum chamber. Critical control of the thick- 
ness of each layer is obtained by the use of a photoelectric monitoring 
system that indicates the thickness of the material as it condenses on 
the glass. Successful mirrors have been made with the coatings on 
the front surface instead of the rear, but damaging of the multilayer 
films by the sputtering of the arc makes such mirrors less practicable. 
A high-temperature lacquer sprayed and baked on the films protects 
them from mechanical damage. It has been found that the lacquer 
actually improves the optical efficiency of the layers by increasing 
their transmission in the infrared region. 

Although mirrors made in this way are still in the development stage 
the tests made on them have been very encouraging. Measurements 

442 G. J. KOCH 

on the first arc lamp samples showed that the illumination at the gate 
contained more visible light and less total radiation than that ob- 
tained from a standard silvered mirror with light shade of Aklo heat- 
absorbing glass in the beam. Even the most efficient heat-absorbing 
glasses show a gradual decrease in transmission from wavelengths of 
600 to 1000 millimicrons; hence they allow an appreciable amount of 
near infrared energy to reach the film. But interference mirrors are 
not limited by this fundamental absorption characteristic. Both 
calculations and experiments have shown that considerably sharper 
cuts can be obtained with interference reflectors than with silvered 
reflectors used with heat-absorbing glasses. In the ultraviolet region 
also, the decrease in reflection with wavelength is sharp but detailed 
measurements have not yet been made. Certainly the peak at 390 
millimicrons in the arc emission curve is greatly reduced. 

A second advantage of such mirrors is that they eliminate the prob- 
lem of breakage of heat-glass. Since the interference films absorb 
little radiation, and since they are distributed over the large area of 
the mirror blank, they do not get nearly so hot as a heat-absorbing 
glass. Actually, the absorption of infrared by the glass mirror blank 
itself is largely responsible for the temperature rise observed. 

Probably the most important advantage these mirrors have over a 
silvered mirror used with a heat-absorbing glass is the control that 
can be attained over the color quality of the light. By proper adjust- 
ment of the thickness of the interference layers, the color of the re- 
flected light can be varied over a wide range. This factor is of major 
importance for the projection of color film. Our experimental arc 
mirrors for 16-mm projectors are coated so that light from a high- 
intensity arc is modified to give satisfactory color balance for the 
projection of Kodachrome film. 

The development work on this type of mirror is still in the labora- 
tory stage. Further details, such as life tests and more efficient com- 
binations of layers, are being investigated. When these tests are 
complete we plan to try the coatings on 14-in. mirrors for tests in 35- 
mm arc projectors. 

Engineering Committees 

New Release Print Leader 

In July, we prematurely announced the availability of a newly proposed release 
print leader. Contrary to expectations, the first prints were found unsatisfac- 
tory after trial at a few of the New York television stations; consequently, an- 
other version was made up and prints were distributed to ah 1 of the television net- 
works in New York. On Friday, September 8, Charles Townsend's subcom- 
mittee met and approved this version for limited trial distribution. After con- 
sultation with Dr. Garman, Chairman of the parent committee, and Fred Bow- 
ditch, Engineering Vice-President, it was agreed to make up a fairly substantial 
quantity of both negatives and prints. All of those requesting samples as a re- 
sult of the July announcement, as well as others known to have an interest in this 
project, have now been supplied. 

There has been some concern among those responsible for this work that rather 
general distribution at this time may cause difficulties. It has been the hope 
from the start that one leader can be developed, which will serve the theater, tele- 
vision and the 16-mm fields equally as well. If the trial leaders now being dis- 
tributed should get into general use, it would certainly mean the existence of two 
leaders and the possibility of considerable confusion; therefore, a letter was sent 
with each sample, requesting the recipients to use the leaders only, for experi- 
mental purposes and to withhold any wider use until the committee has had an 
opportunity to review all comments. 

Committee Ballots 

During the last six months, members of Engineering Committees have been ex- 
tremely lax about returning their ballots on proposed standards, recommenda- 
tions or committee reports. Often when members feel that they have no com- 
ments or critical interest in a particular subject, they fail to take any action to 
notify Society headquarters to that effect. Consequently, the balloting is not 
closed because of the belief that the missing votes represent objections to a pro- 
posal. It is realized that in some cases committee members need time to make a 
thorough investigation before casting a formal ballot. In those cases, however, 
it would be most helpful if such were reported so as to save the time and money 
spent sending out repeated follow-ups. 

At the present time, six engineering ballots are in process. One, " 16-mm Re- 
view Room Characteristics," was mailed on May 29, 1950. The ballot was sent 
to 72 committee members and to date just half have been returned. This is a 
very poor response for a project that two years ago was considered important 
enough to warrant considerable activity. Listening tests and meetings were held 
in Astoria, New York City, Washington and Chicago. Equipment was shipped 
around arid committee members spent time and money traveling to those ses- 
sions. After all of that time and effort, it is most unfortunate that the project 
should be held up merely because members neglect to vote. 

Possibly, there is some fault in our method of operation which is causing the 
delays. If this is true, we would greatly appreciate your suggestions. 


Book Reviews 

Questions and Answers in Television Engineering, by Carter V. 

Rabinoff and Magdalena E. Walbrecht 

Published (1950) by McGraw-Hill, 330 West 42d St., New York 18. 283 pp. + 
2 pp. appendix + 16 pp. index. 175 illus. 6 X 9 in. Price $4.50. 

Although book lists today abound with volumes whose titles include "tele- 
vision," by far the majority deal with television receiver servicing or with the 
theory and design of receivers. Books which treat television as a many-faceted 
field, no part of which is more important than the other parts, are not common. 
This volume is such a work. 

It is divided into twelve chapters: Antennas, Radiation, and Transmission 
Lines; R.F. and I.F. Amplifiers, Converters, and Oscillators; The Limiter, 
Clipper, Detector, and Sync Separator; Video Amplifiers; Deflection Systems; 
Cathode-Ray Tubes, Camera Tubes, and Photoelectric Cells; Images and Pat- 
terns; Optical Systems and Illumination; Transmitters; Standards, Laws, and 
Regulations; Receiver Principles, Filter Theory, and Power Supplies; and An- 
alysis of Two Typical Receivers. 

The book is intended as a reference source for those who are in the television 
field and as a "semi-textbook" for those who need to clarify and organize their 
knowledge. It is singularly successful in both roles, primarily because of the adop- 
tion and intelligent treatment of the question-and-answer technique. A writer 
whose object is simply to give an explanation of a certain rather wide subject too 
often tends to ramble, to lack organization, to use too many words, and to avoid 
being specific. 

The authors here, however, have voluntarily bound themselves to the asking ol 
the logical questions and to answering them specifically, succinctly and pre- 
cisely. In this way, they have made each piece of information definitely useful 
in doing a certain job or in the study of a particular phenomenon or piece of equip- 
ment. They have paralleled the thoughts of the reader who uses the book for 
reference, for his reason for referring to the book in the first place is because he has 
a question in mind. 

As a learning aid for readers who need to consolidate their knowledge the book 
is valuable because the questions have been chosen to carry the subject through 
in logical order. The reader who starts at the beginning of the book need not be 
an electronics specialist. He is introduced to the special character of v.h.f. 
propagation, the reasons for antenna design and the practical calculations, re- 
ceiver inputs and other receiver circuits. The explanation of scanning methods 
leads into a discussion of cathode-ray tubes and photocells, which naturally carries 
forward to images and patterns. These three subjects are all linked with both 
transmission and reception. The next section, on optical systems and illumina- 
tion, will be quickly understood by motion picture men and will make the pieces 
of the previous explanations fall into place neatly. 

The section on transmitters is basic enough for instruction, but is also specific 
enough, with elementary calculations and figures to be of aid to the television 
specialist. Many of these questions, in fact, parallel those in FCC licence ex- 
aminations, and may be used as review material for prospective examinees. In 
this section there is particular reference to the special applications of motion pic- 
ture techniques to television. 

The important definitions, FCC regulations, and standard practices in trans- 
mission are given in the standards section. The receiver section is devoted largely 


to the points that do not ordinarily get full treatment, such as projection systems, 
filter configurations, and power supplies. The final chapter is a stage-by-stage 
breakdown of an RCA and a GE receiver from a design standpoint, with, in many 
questions, the interesting and instructive approach of asking the "why" of certain 
design provisions. The appendix contains additional information on proposed 
u.h.f. channels, metropolitan and community stations, interference ratios, allo- 
cations, power radiation regulations, and auxiliary (mobile pickup and relay) 
stations. The index is unusually complete and helpful. 

This volume is not large enough to contain all the information that every reader 
might need at some time or other; but it is certainly a highly useful one to keep 
in any technical library, especially in conjunction with some of the more con- 
ventional books on receivers and other specialized divisions of the television 
field. RICHARD H. DORF, Television Consultant, 255 W. 84th St., New York 24. 

Reunions D'Opticiens, Tenues a Paris en Octobre 1946, Textes ras- 
sembles par Pierre Fleury, Andre Marechal et Mme. Claire Ang- 

lade, Institut d'Optique, Paris 

Published (1950) by Editions de la Revue d'Optique, 165, Rue de Sevres; 3 et 5, 
Boulevard Pasteur, Paris (15 e ). xv + 673 pp. 6X9 in. Price on request. 

This is a collection of 131 scientific papers presented by 109 authors at an inter- 
national convention held in Paris in October, 1946. The volume contains numer- 
ous illustrations. It represents a summary of all modern research trends in the 
field of theoretical and applied optics. 

The volume covers the following topics: basic theoretical studies, theory of 
aberrations, structure and perception of optical images, optical instruments, 
optical surfaces and materials, optical measurements, sources and receptors 
physiological optics and color, ophthalmological optics, spectroscopy, molecular 
optics, polarimetry, astronomical and atmospherical optics, and corpuscular 

While being of much interest to anyone interested in fundamental optical in- 
vestigations, this volume is not intended as a textbook nor can it be used as such; 
and, of course, it is not an engineering manual. For this reason, a practical mo- 
tion picture engineer will not find in this volume much that is of direct use to him, 
unless he wants to make himself acquainted with powerful new ideas and tools 
which at some future time may considerably influence motion picture art and 
engineering. DR. K. PESTRECOV, Bausch & Lomb Optical Co., Rochester 2, 

Photographic Instantanee et Cinematographic Ultra-Rapide, par 

P. Fayolle et P. Naslin 

Published (1950) by Editions de la Revue d'Optique, 165, Rue de Sevres; 3 et 5, 
Boulevard Pasteur, Paris ( 15 e ). Price on request. 

Messrs. Fayolle and Naslin have undertaken to produce a text on high-speed 
photography, something that has been needed for a long time. There have been 
previous publications but they have generally covered only a small portion of the 
field. Tulpholme in his book, Photography in Engineering, discusses high-speed 
photography in one chapter. Other phases have been covered hi Lehrbuch der 
Ballistick and in other numerous papers. Now, Messrs. Fayolle and Naslin have 
consolidated much of this lore in one text. 

The summarization that they have given covering the optical problems of high- 


speed photography is well done. The discussions of light sources and the various 
cameras which are in existence are good; and, for a text, to serve as a background 
in high-speed photography, it is invaluable today, if for no other reason than that 
no other such source exists. 

There are, however, some exceptions to be taken to the text and your reviewer 
has written the authors that they have neglected many of the advances that have 
been made in the United States in the last two or three years, advances which have 
been described in the symposia on high-speed photography held by the Society 
of Motion Picture and Television Engineers. Such include the high pulse-rate 
Edgerton flash unit, the late advances in flashtube design by General Electric, the 
design of rotating prism cameras, and the development and use of high-speed 
motion picture cameras in this country. Furthermore, there is no reference to 
the incandescent lights which have been developed by General Electric for high- 
speed photography. Your reviewer might be a little biased but he feels that, if 
rotating prism cameras are discussed, then the two kinds that are manufactured in 
this country should be discussed rather than confining the description to one. 
Since the two cameras are about equally popular in this country, an impartial 
review should include discussions of both. 

For those who are seriously interested in high-speed photography, it is felt that 
this book is a must. Ability to read French is not a requirement for understand- 
ing this book quite thoroughly. 

It should be noted that the next major project of your Society's High-Speed 
Photography Commitlee is to produce an American textbook. Whether that 
will be chiefly the work of one man or the result of the efforts of the Committee 
as a whole is not yet established, but there will soon be a text in English in pro- 
duction. JOHN H. WADDELL, Industrial and Technical Photographic Division 
Fastax Cameras, Wollensak Optical Co., Rochester 21, N. Y. 

Letters to the Editor 

Re: Light Measurement for Exposure Control 

[The publication of a paper under this title by Don Norwood, May 1950 JOURNAL 

pp. 585-602, has elicited the following comments. ] 

I was very interested to read the above noted article, for my development of the 
Duplex Incident Light Exposure Meter Technique has meant that for the last 
three years or so I have been working on almost parallel lines. 

Before noting some further developments, I hope you will permit me to correct 
Mr. Norwood's statement on p. 595, in which he claims that his "concept of 
Effective Illumination, which takes into account illumination intensity and rela- 
tive positions of observer, subject and light source has not heretofore been crystallized 
or formulated" [the latter italics are mine]. In contradiction to this statement, 
reference may be made to the November 1948 British Kinematography containing 
my article "Exposure Technique for Reversal Materials," from which it may be 
noted that I described this effect (which I named the "Duplex" technique) over 
two years ago before a meeting of the British Kinematograph Society in London. 

Since reading my paper in 1948, a slight improvement over my originally 
described "Horizontal Duplex" method has been developed. The new recom- 
mended "Direct Duplex" method, which has been worked out by my colleague 
L. C. Walshe and myself, still consists of taking two readings with an incident 
light meter (of the flat window type), namely a "camera direction" reading (as 


previously recommended) and a "major source direction" reading (as originally 
recommended for use alone by P. C. Smethurst, who first introduced the incident 
light meter technique in England in 1936 see his paper in British Kinematog- 
raphy,vol. l,no. 1). 

The required exposure for average work is then given by simply taking the 
geometric mean of the two Duplex readings, i.e., the mid-point on the stop scale 
between the two readings. For clear-cut conditions this technique will be found 
to line up almost exactly with the exposure levels recommended by the principal 
reversal color film manufacturers, and it has proved to be highly successful in 
practice. It has, incidentally, already been fully described in a book on this sub- 
ject which is in preparation and will be published in due course .... 

July 21, 1950 1 Deneway, Bramhall 

Cheshire, England 

In comparison with Mr. Dunn's statement that he has been working on almost 
parallel lines for the last three years or so, I should note that my experience with 
incident light exposure meters date from 1933 when my first incident light ex- 
posure meter (of the hemisphere light-collector type) was constructed. That 
meter was rigorously tested for several years, and patent application was made 
in 1938 (U.S. Patent 2,214,283). 

The principle involved in that meter took into precise account both light in- 
tensity and the geometric relationship of keylight, subject and camera, as de- 
scribed in the aforementioned paper. 

Work on other types of incident light exposure meters, which are basad on the 
same fundamental principle, has proceeded since that time. See: U. 3. Pat. 
#2,489,664, application filed in 1946; and U.S. Pat. #2,444,464, application filed in 
1947. Both applications were filed prior to the time when Mr. Dunn reports the 
beginning of his endeavors, circa 1947. Pkplanation was made on p. 585 of May 
1950 JOURNAL, that evolving patent protection had made full release of basic data 
inadvisable until 1948, when the paper was written. The JOURNAL publication 
showed that the Society received the paper in February 1949. 

It has been recognized that some workers in this field have had a more or less 
hazy realization that more was involved in incident light measurement for ex- 
posure control than a simple measurement of light intensity. Various corrective 
expedients have been proposed by some of these workers, such as pointing a meter 
with a plane surface light collector toward the camera from subject's position; 
pointing said meter toward principal light source; aiming said meter toward a 
point halfway between said light source and said camera; pointing meter at 
camera and at light source in turn and using a mean reading as suggested by Mr. 
Dunn. However, none of these makeshift methods appears to indicate a full and 
clear-cut realization of the basic principles involved in the matter. None of the 
experimenters have, to my knowledge, brought forth precise and comprehensive 
formulae such as those shown in (15) and (16) on p. 595 of the May JOURNAL. 

I do not agree that Mr. Dunn in describing his Duplex Method in British 
Kinematography has given a clear-cut, well crystallized comprehension of ah! the 
factors involved, as well as a formula for accurate solution of the problem. For 
instance, his formula for calibration of incident light meters was preceded by a 
quite similar formula on p. 14-6 of the I.E.S. Lighting Handbook, published in 
1947. Neither formula takes into account the vital factor of geometrical rela- 
tionship of subject, camera and light source. If this relationship were under- 
stood it would seem that it would have been put into Mr. Dunn's formula. 


It is of interest to examine under specific conditions Mr. Dunn's recommended 
method of operation. As an example, we find that Mr. Dunn's Direct Duplex 
Method would give identical exposure control settings for a cross-lighted arrange- 
ment (90 keylight angle), and a back-lighted arrangement (135 keylight angle), 
where other factors remain constant. Reference to instructions issued by leading 
color film manufacturers (see Note 5, my paper) will show that a considerable 
difference in exposure control setting is recommended for cross-lighted and back- 
lighted conditions. The difference is usually about one /-stop. It is generally 
believed that reversal color film will not tolerate an error of one /-stop. There- 
fore Mr. Dunn's "two-readings" method when used with color film under the com- 
mon conditions described above will give errors which lie outside the acceptance 
latitude of the film. This also negates Mr. Dunn's statement (next to last para- 
graph in his letter) that his method lines up with recommendations of film manu- 

August 5, 1950 1470 San Pasqual St. 

Pasadena 5, Calif. 

New Members 

The following have been added to the Society's rolls since the list published last month. 
The designations of grades are the same as those in the 1950 MEMBERSHIP DIRECTORY: 

Honorary (H) Fellow (F) Active (M) Associate (A) Student (S) 

Alaimo. James J., American Television 
Inst. Mail: 4652 N. Kenmore Ave., 
Chicago 40, 111. (S) 

Anderson, A. Stephen, Recording Techni- 
cian, 949 Third Ave., New York 22, 
N.Y. (M) 

Band, Edward A., Television Engineer, 
National Broadcasting Co. Mail: 1317 
Second Ave., New York 21, N.Y. (A) 

Benham, E. E., Television Engineer, 
KTTV, Inc. Mail: 5240 Beeman, 
North Hollywood, Calif. (A) 

Blackwell, L. H., Cinematograph Engi- 
neer, L. H. Blackwell & Co., 133 Em- 
pire Rd., Perivale, Greenford, Middle- 
sex, England. (M) 

Brown, Robert J., American Television 
Inst. Mail: Arlington Trailer Ct., 
Arlington Heights, 111. (S) 

Buckingham, William D., Engineer, 
Western Union Telegraph Co. Mail: 
86 Post Lane, Southampton, Long 
Island, N.Y. (M) 

Casterlin, Charles C., Head Paying-Re- 
ceiving Teller, Hamilton National 
Bank, Dupont Circle Branch. Mail: 
3345 Stephenson PI., N.W., Washing- 
ton 15, D.C. (A) 

Chinn, Howard A., Chief Audio & Video 
Engineer, Columbia Broadcasting Sys- 
tem, Inc. Mail: 6 Knollwood Rd., 
Tuckahoe, N.Y. (M) 

Creutz, John, Radio Engineer, Old 
Dominion Dr., McLean, Va. (A) 


Davies, Hugh B., Sound Service Techni- 
cian, Gaumont-Kalee Ltd. Mail: 6 
Cornish Rd., Toronto, Ontario, Can- 
ada. (A) 

Davis, Harold C., American Television 
Inst. Mail: 1313 W. Grenshaw, Chi- 
cago, 111. (S) 

Diebold, Jerome C., Executive Assistant, 
Wilding Picture Productions, Inc. 
Mail: 1345 Argyle St., Chicago, 111. 

du Preez, H. S. J., Technical Organizer, 
Karfo Films. Mail: 53 Sixth Ave., 
Melville, Johannesburg, South Africa. 

Eddey, Erwin, Sound Timer, DeLuxe 
Laboratories, Inc. Mail: 15 Lindley 
Ave., Tenafly, N.J. (A) 

Faust, Roland J., Motion Picture Writer- 
Director, Indiana University. Mail: 
2126 E. Seventh St., Bloomington, 
Ind. (M) 

Fegan, Albert A., Electronic Technician- 
Projectionist, U. S. Navy & Local 
Theaters. Mail: 17 W. Magnolia St., 
Stockton 3, Calif. (A) 

Foulds, Blair, Commercial Engineering 
Manager, General Precision Lab., Inc., 
63 Bedford Rd., Pleasantville, N.Y. 

Gawel, Eugene W., American Television 
Inst. Mail: 8200 Brandon Ave., Chi- 
cago 17, 111. (S) 

Giroux, George R., Jr., Television Assis- 
tant Director, KTTV, Inc. Mail: 800 
N. Maple, Burbank, Calif. (A) 

Hicks, Willard L., Head, Engineering 
Photographic Dept., Burroughs Adding 
Machine Co. Mail: 4114 Larchmont 
Ave., Detroit 4, Mich. (M) 

Horn, James H., Sales, Ampro Corp. 
Mail: 140 Riverside Dr., New York 24, 
N.Y. (A) 

Jensen, Stanley E., Quality Control, Mo- 
tion Picture Processing, Eastman 
Kodak Co. Mail: 3319 Greenbrier 
Dr., Dallas, Tex. (A) 

Johnson, James, Sound Engineer, British 
Lion Studios. Mail: 18 Colbrook 
Ave., Hayes, Middlesex, England. (M) 

Johnson, Warner A., Development Engi- 
neer, Minneapolis Honeywell Regula- 
tor Co. Mail: 1322 S. Chicago, Free- 
port, 111. (A) 

Johnston, Kenneth S., Theater Sound In- 
spector, Gaumont-Kalee Ltd. Mail: 
140 Sunnyside Ave., Toronto, Ontario, 
Canada. (A) 

Kay, William, Assistant Manager, 16- 
Mm. Motion Picture Exchange, Film 
Center, Inc. Mail: 3606 New England 
Ave., Chicago, 111. (A) 

Kreuser, John A., Camera Designer, Bell 
& Howell Co. Mail: 3015 N. Nagle 
Ave., Chicago 34, 111. (A) 

Lefkowitz, George, RCA Institute. Mail: 
592 Beech Terrace, Bronx 54, N.Y. (S) 

Lewis, Lawrence, Projectionist, 20th 
Century Theatres. Mail: 8 Rossmore 
Rd., Toronto, Ontario, Canada. (A) 

McGill, Howard L, Sales & Engineering, 
Zack Radio Supply Co. Mail: 1426 
Market St., San Francisco, Calif. (A) 

McNamee, Raymond J., American Tele- 
vision Inst. Mail: 4915 W. Lexington 
St., Chicago 44, 111. (S) 

Mitchell, Wayne, University of Southern 
California. Mail: 3520 South Hoover 
Blvd., Los Angeles 7, Calif. (S) 

Morrow, Donald J., Cameraman-Direc- 
tor, Educational Film Lab., c/o U. S. 
Indian School, Santa Fe, N.M. (A) 

Nelson, Charles R., American Television 
Inst. Mail: 4906 N. Kenmore Ave., 
Chicago, 111. (S) 

Pacent, Louis G., President, Pacent Engi- 
neering Corp., 79 Madison Ave., New 
York 16, N.Y. (F) 

Pike, Donald E., Technical Director, Na- 
tional Broadcasting Co. Mail: 87 
Garrabrant Ave., Bloomfield, N.J. (A) 

Porrett, Fred, Motion Picture Camera- 
man, 106 Washington Place, New York 
14, N.Y. (M) 

Rouzer, Danny, Cameraman, Cine-Tele 
Productions. Mail: 7022 Melrose 
Ave., Hollywood 28, Calif. (A) 

Scho field, Henry J., Assistant Camera- 
man and Laboratory Technician, D'- 

Andrea Films. Mail: 2511 West End 
Ave., Nashville, Tenn. (A) 

Schuchman, Nathan, Television Camera- 
man, NBC-TV. Mail: 2137 Ocean 
Parkway, Brooklyn 23, N.Y. (A) 

Seeger, Edward W., American Television 
Inst. Mail: 418 Franklin Ave., River 
Forest, 111. (S) 

Sidel, H. P., School Director and Writer, 
The School of Modern Photography. 
Mail: 410 E. 57 St., New York 22, 
N.Y. (M) 

Silverman, Samuel L., Purchasing Agent 
and Credit Manager, Precision Film 
Laboratories, Inc. Mail: 1844 E. 
Fourth St., Brooklyn 23, N.Y. (M) 

Smith, Fred S., American Television Inst. 
Mail: 4520 N. Hazel, Chicago 40, 111. 

Soltys, Richard J., University of Southern 
California. Mail: 1114 North Cedar 
St., Glendale 7, Calif. (S) 

Somers, Arthur A., Photographer, 3412 
Superior Park Dr., Cleveland Heights, 
Ohio. (A) 

Stedman, Dean C., University of South- 
ern California. Mail: P.O. Box 766, 
Cluster, S.D. (S) 

Swift, William C. G., Sound Engineer, 
Paramount Pictures Corp. Mail: 346 
Richbell Rd., Mamaroneck, N.Y. (A) 

Taraba, Vilem, Technical Manager, State 
Film Studios, Barrandov, Prague, 
Czechoslovakia. (A) 

Young, Charles W., Photographer, Penn- 
sylvania State College Ordnance Re- 
search Laboratory. Mail: 643 Fair- 
way Rd., State College, Pa. (A) 

Young, Keith D., University of Southern 
California. Mail: Willard Hall, 942 W. 
34 St., Los Angeles 7, Calif. (S) 

Herold, Ralph E., Audio-Visual Coordi- 
nator, Los Angeles Public Library. 
Mail: 5069 Montezuma St., Los Ange- 
les 42, Calif. (A) to (M) 
Morrison, Ernest W., Cameraman, U.S. 
Air Force, Lookout Maintenance Lab- 
oratory. Mail: 564 Larchmont Blvd.. 
Los Angeles 4, Calif. (S) to (A) 
Thomas, John W., Radio and Sound 
Serviceman, 3736 South Michigan 
Ave., Chicago, 111. (S) to (A) 


Haines, Robert A., because of a copy- 
editing error, appears in the 1950 
Membership Directory with an address 
(at St. Louis) which is out of date since 
1946. The listing should be: 
Haines, Robert A., Senior Engineer, 
FEC Motion Picture Service, SSS, 
GHQ, Far East Command, APO 500, 
c/o Postmaster, San Francisco, Calif. 


New Products 

Further information concerning the material described below can be 
obtained by writing direct to the manufacturers. As in the case of 
technical papers, publication of these news items does not constitute 
endorsement of the manufacturer's statements nor of his products. 

All-metal reflectors, made by combining two metals by electro- 
plating, have a front surface protective coating of rhodium 
which is an element from the platinum family, exceptionally 
high in reflectivity and reported the only metal available with 
sufficient reflectance and able to withstand successfully the 
damaging effects of high-intensity carbon arcs. . The manu- 
facturer, Heyer-Shultz, Inc., Cedar Grove, N.J., reports that 
fabrication to close tolerances permits interchanging similar 
models made years apart, with a minimum of reflector adjust- 
ments by the projectionist. 

PM Hysteresis clutches and brakes were recently developed by Duncan & 
Bailey, Inc., 785 Hertel Ave., Buffalo, N.Y. They provide for smooth applica- 
tion of torque from conditions of 100% slip to zero slip or operate continuously at 
any slip or torque condition through a wide range of rotational speeds and power 
capacities. Bearing drag is the only load present when these devices operate in 
the unenergized conditions and the characteristics of power application depend 
upon the performance of the control mechanisms employed. 

Torque bears a linear relation to control current and for instantaneous applica- 
tion of control power engagement is rapid but shock loads are absent and response 
time is generally proportional to control power impressed. A typical value for 
30 w of applied control power and 100 oz-in. of torque is 100 milliseconds. A one- 
quarter horsepower unit operating at 1750 rpm has frequency response of 2 cycles 
per second. 

Control current required is roughly one-tenth of the power output of the device 
when operated as a coupling. Type PM-3 using 25 w of control power will trans- 
mit 375 w at 3400 rpm. 

With any of a wide variety of servo or external control systems, these units 
working as either transmitters or drag elements can be used to provide constant 
speed, torque, tension or repetitive cycling within performance limits of about 


1%. Input rotational speeds up to 10,000 cycles are permissible as are rapid 
cycling operations, limited only by resulting rise in temperature. Design of the 
housing provides for sufficient heat dissipation and performance is within rated 
capacities as long as temperature rise does not exceed 300 F for continuous oper- 

Illustrations show the external clutch housing and one constant tension ap- 
plication employing two units one as a transmitter and one as a brake, actuated 
by a sensing device which responds to changes in diameter of the loaded feed 
spool. Detailed performance data and dimensions are available from the manu- 
facturer who also designed the FM (fluid magnetic) clutch for specialized and 
somewhat less severe applications. Magnetic friction grab couplings for indus- 
trial applications with torque ratings of 80 in.-oz or 40 ft-lb are also available. 




A special viewfinder 
ground glass for 35-mm 
motion picture cameras 
shows the portion of the 
scene most likely to be 
reproduced on a home 
television receiver. This 
pattern gives the cinema- 
tographer an accurate 
check of the average cut- 
off losses which must be 
allowed for in order to 
have a good visual pres- 
entation on television. 
The etched pattern is 
based on the first section 

of the Society of Motion Picture and Television Engineers test film as described 
and illustrated in the February 1950, JOURNAL, p. 211. The striped area rep- 
resents the section lost at the time of scanning in the television station. The in- 
side rectangle is approximately 80% of the picture area. Tests indicate that this 
area is reasonably well reproduced on most home receivers. Television Ground 
Glass for the Mitchell camera etched with the television aperture is available 
directly from Greiner Glass Industries Co., 781 E. 142 St., New York 54. Tele- 
vision Ground Glass for other cameras is also available. Prices on request. 

Ground Glass X 3 

SMPTE Officers and Committees: The Roster of Society Officers was 
published in the May JOURNAL. For Administrative Committees see 
pp. 515-518 of the April 1950 JOURNAL. The most recent roster of 
Engineering Committees is on pp. 337-340 of September 1950 JOURNAL. 

Meetings of Other Societies 

Audio Engineering Society, National Convention, Oct. 26-28, Hotel New Yorker, 

New York 

Optical Society of America, Oct. 26-28, New York 

Theatre Owners of America, Annual Convention, Oct. 30-Nov. 2, Shamrock 

Hotel, Houston, Texas 
Acoustical Society of America, Fall Meeting, Nov. 9-11, Boston 


Statement of the Ownership, Management, Circulation, Etc., Required by the 
Act of Congress of August 24, 1912, as Amended by the Acts of March 3, 1933, 
and July 2, 1946, of Journal of the Society of Motion Picture and Television Engi- 
neers, published monthly at Easton, Pa., for October 1, 1950. 

State of New York ) 
County of New York [ s ^ 

Before me, a Notary Public in and for the State and county aforesaid, person- 
ally appeared Boyce Nemec, who, having been duly sworn according to law, de- 
poses and says that he is the Executive Secretary of the Journal of the Society of 
Motion Picture and Television Engineers and that the following is, to the best of 
his knowledge and belief, a true statement of the ownership, management (and 
if a daily, weekly, semiweekly, or triweekly newspaper, the circulation), etc., of the 
aforesaid publication for the date shown in the above caption, required by the Act 
of August 24, 1912, as amended by the Acts of March 3, 1933, and July 2, 1946, 
embodied in section 537, Postal Laws and Regulations, printed on the reverse of 
this form, to wit: 

1. That the names and addresses of the publisher, editor, managing editor, 
and business managers are: 

Name of Post Office Address 

Publisher, Society of Motion Picture and Television Engineers, Inc., 342 Madi- 
son Ave., New York 17, N. Y. 

Editor, Victor H. Allen, 342 Madison Ave., New York 17, N. Y. 
Managing Editor, None. 
Business Manager, Boyce Nemec, 342 Madison Ave., New York 17, N. Y. 

2. That the owner is: (If owned by a corporation, its name and address 
must be stated and also immediately thereunder the names and addresses of 
stockholders owning or holding one per cent or more of total amount of stock. 
If not owned by a corporation, the names and addresses of the individual owners 
must be given. If owned by a firm, company, or other unincorporated concern, 
its name and address, as well as those of each individual member, must be given.) 
Society of Motion Picture and Television Engineers, Inc., 342 Madison Ave., New 
York 17, N. Y. 

Earl I. Sponable, President, 460 W. 54 St., New York 19, N. Y. 
Robert M. Corbin, Secretary, 343 State St., Rochester 4, N. Y. 
Frank E. Cahill, Jr., Treasurer, 321 W. 44 St., New York 18, N. Y. 
No stockholders. 

3. That the known bondholders, mortgagees, and other security holders 
owning or holding one per cent or more of total amount of bonds, mortgages, or 
other securities are: (If there are none, so state.) 


4. That the two paragraphs next above, giving the names of the owners, 
stockholders, and security holders, if any, contain not only the list of stockholders 
and security holders as they appear upon the books of the company but also, 
in cases where the stockholder or security holder appears upon the books of the 
company as trustee or in any other fiduciary relation, the name of the person or 
corporation for whom such trustee is acting, is given; also that the said two 
paragraphs contain statements embracing affiant's full knowledge and belief 
as to the circumstances and conditions under which stockholders and security 
holders who do not appear upon the books of the company as trustees, hold stock 
and securities in a capacity other than that of a bona fide owner ; and this affiant 
has no reason to believe that any other person, association, or corporation has 
any interest direct or indirect in the said stock, bonds, or other securities than 
as so stated by him. 

5. That the average number of copies of each issue of this publication sold 
or distributed, through the mails or otherwise, to paid subscribers during the 
twelve months preceding the date shown above is: (This information is required 
from daily, weekly, semiweekly, and triweekly newspapers only.) 

BOYCE NEMEC, Exec. Secy., Business Manager. 
Sworn to and subscribed before me this 21st day of September, 1950. 

(Seal) Elisabeth J. Rubino 
Notary Public, No. 41-3390800, Queens 

(My commission expires March 30, 1951) 

Journal of the Society of 

Motion Picture and Television Engineers 


Synthetic Color-Forming Binders for Photographic Emulsions .... 


An Improved Video System for Television Studios . NEWLAND F. SMITH 477 

Infrared Photography with Electric-Flash . . FREDERICK E. BARSTOW 485 

Magnetic Sound Film Developments in Great Britain . 0. K. KOLB 496 

Improvements in Large-Screen Television Projection . T. M. C. LANCE 509 

Trends of 16-Mm Projector Equipment in the Army . JAMES A. MOSES 525 

Foreign Versions VICTOR VOLMAR 536 

A Progress Report of Engineering Committee Work . F. T. BOWDITCH 547 

Biological Photographic Association 549 

Current Literature 550 

New Members 551 


Photographic Optics, by Allen R. Greenleaf 

Reviewed by Oscar W. Richards 552 

A Grammar of the Film, by Raymond Spottiswoode 

Reviewed by Russell C. Holslag 553 

New Products 554 


Chairman Editor Chairman 

Board of Editors Papers Committee 

Subscription to nonmembers, $12.50 per annum; to members, $6.25 per annum, included in 
their annual membership dues; single copies, $1.50. Order from the Society's General Office. 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and 
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture and Tele- 
vision Engineers, Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General 
and Editorial Office, 342 Madison Ave., New York 17, N.Y. Entered as second-class matter 
January 15, 1930, at the Post Office at Easton, Pa., under the Act of March 3, 1879. 
Copyright, 1950, by the Society of Motion Picture and Television Engineers, Inc. Permission 
to republish material from the JOURNAL must be obtained in writing from the General Office of 
the Society. Copyright under International Copyright Convention and Pan-American Con- 
vention. The Society is not responsible for statements of authors or contributors. 

Society of 

Motion Picture and Television Engineers 

342 MADISON AVENUE, NEW YORK 17, N.Y., TEL. Mu 2-2185 

Earl I. Sponable 
460 W. 54 St. 
New York 19, N.Y. 

Peter Mole 

941 N. Sycamore Ave. 
Hollywood 38, Calif. 



Loren L. Ryder 
5451 Marathon St. 
Hollywood 38, Calif. 

Clyde R. Keith 
120 Broadway 
New York 5, N.Y. 

Robert M. Corbin 
343 State St. 
Rochester 4, N.Y. 

William C. Kunzmann 
Box 6087 
Cleveland 1, Ohio 

Fred T. Bowditch 
Box 6087 
Cleveland 1, Ohio 


Frank E. Cahill, Jr. 
321 W. 44 St. 
New York 18, N.Y. 

Ralph B. Austrian 
25 W. 54 St. 
New York 19, N.Y. 


Herbert Barnett 
Manville Lane 
Pleasantville, N.Y. 

Kenneth F. Morgan 
6601 Romaine St. 
Los Angeles 38, Calif. 

Norwood L. Simmons 
6706 Santa Monica Blvd. 
Hollywood 38, Calif. 


F. E. Carlson 
Nela Park 
Cleveland 12, Ohio 

George W. Colburn 
164 N. Wacker Dr. 
Chicago 6, HI. 


Charles R. Daily 
113 N. Laurel Ave. 
Los Angeles 36, Calif. 

John P. Livadary 
4034 Cromwell Ave. 
Los Angeles, Calif. 

William B. Lodge 
485 Madison Ave. 
New York 22, N.Y. 

Edmund A. Bertram 
850 Tenth Ave. 
New York 19, N.Y. 

Malcolm G. Townsley 
7100 McCormick Rd. 
Chicago 45, 111. 


Lorin D. Grignon 
1427 Warnall Ave. 
Los Angeles 24, Calif. 

Paul J. Larsen 
4313 Center St. 
Chevy Chase, Md. 

William H. Rivers 
342 Madison Ave. 
New York 17, N.Y. 

Edward S. Seeley 
161 Sixth Ave. 
New York 13, N.Y. 

R. T. Van Niman 
4441 Indianola Ave. 
Indianapolis, Ind. 

Synthetic Color-Forming Binders 
For Photographic Emulsions 



SUMMARY: The development of synthetic color-forming binders and 
their application to photographic emulsions is discussed. These accomplish- 
ments have made possible the manufacture of a release positive color film 
designated Du Pont Type 275. * A resume of some of the novel features of the 
stock is given and the utilization of the material as a color release medium is 
covered. Details of the printing and processing of both picture and sound 
records are given. 

ONE OF THE WAYS of creating a dye image in proportion to the 
latent image in a photographic film is known as coupling color de- 
velopment, t It was disclosed originally by Rudolph Fischer 1 in 1912. 
In fact, in 1914, Fischer and Siegrist 2 published the results of a thorough 
study of the chemistry involved, and disclosed several classes of reac- 
tions that could be used and suggested various photographic elements 
utilizing them. Of these classes of reactions, those useful in the for- 
mation of the azomethine and indophenolt dyes involved in the 
process to be described later can be represented by the following 

,+H 2 G' +4AG+ 



^ +4H+ 


PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. 

* The product on safety base is now designated Du Pont Type 875. 

f It has also been called indirect color development, secondary color development, 

dye coupling development, and color-forming development. 
| Properly called indoaniline dyes by the strictest chemical authorities, but almost 

always referred to as indophenol dyes in the photographic literature. 



Although the exact mechanism of these reactions is still not com- 
pletely understood, an oxidation product of the developing agent pro- 
duced in the reduction of the exposed silver halide reacts in some way 
with the coupler to form the dye in direct proportion to the amount of 
silver formed. When the silver is removed, only the dye image re- 

In addition to comprehending the full possibilities of coupling color 
development, Fischer and Siegrist disclosed a broad picture of the 
various specific classes of developing agents and couplers that were 
applicable. Over the years this knowledge has been extended by 
numerous investigators. 


Today there are commercial processes involving the application of 
the above reactions, for example, those in which the couplers are in 
developing solutions with the developing agent and those in which the 
couplers are in the emulsions. The former yields monochrome pic- 
tures readily, but three-color pictures only by rather complicated proc- 
essing procedures. The latter readily yields three-color pictures, 
avoiding the cumbersome processing, provided the couplers are im- 
mobilized in their respective layers in the film. Such immobilization 
is necessary to avoid contamination of the various layers through 
wandering of the couplers. 

The first solution to the coupler mobility problem came by placing 
substituents in the color coupler molecules in positions which did not 
affect the coupling power or quality. Although a reasonably high 
solubility in alkaline solutions was retained, these substituents in- 
creased the molecular dimensions of the couplers considerably and 
thus reduced the rate of diffusion from one layer to another to a 
tolerable amount. Such a system thus utilizes three chief components 
in its emulsions: (1) silver halide, (2) binder and (3) coupler. 

The newest method devised for overcoming coupler mobility, and 
one that at the same time offers other advantages, is the use of syn- 
thetic binders for the photographic silver halides which are at the 
same time couplers. Such binders make possible the complete elim- 
ination of gelatin from emulsions useful for color photography. Since 
the color-forming groups are a part of the binder, the use of a third 
component in addition to the gelatin and silver halide of the basic 
black-and-white emulsions is not necessary. 



It will be recalled from experience in the handling of black-and- 
white films that the development of the photographic image, unlike 
many familiar chemical reactions, is not an "instantaneous" process. 
This is because in addition to the time required for diffusion of the 
developer solution throughout the emulsion layer, the act of develop- 
ment itself is regarded as a surface reaction taking place at the inter- 
face of the emulsion grains and the liquid developer solution. 

In coupling color development the situation becomes still more 
complex because a third reacting species, namely, the color coupler, is 
involved and because the developing agent itself, after becoming 
partially oxidized in the development step, must then undergo a 
second reaction with the color coupler. Before this second process can 
occur, however, it becomes necessary for the partially oxidized de- 
veloper to move about in search of coupler molecules with which to 
react. It is because of this sequence of steps in the color coupling 
process that the dye deposits may not reside in the immediate locale 
of the developed grains of silver but rather in a diffuse cloud nearby. 
It has been noted earlier that Fischer's early work in this field had 
been extended with a variety of techniques for making monomeric 
color couplers immobile in emulsion layers. Under such circum- 
stances, however, the bulk of the coupler may be situated at some dis- 
tance from the silver halide grains, making it necessary for migration 
of the intermediate reaction products of the color developer to take 
place. As a consequence, reaction with coupler molecules may occur 
diffusely in the vicinity of the grain, rather than in a concentrated 
zone at the surface of the grain. Furthermore, during the migration of 
the partially oxidized developer molecules, secondary reactions may 
occur, thereby reducing still further the efficiency of the over-all 

The utilization of chemically combined color coupling nuclei in a 
polymer molecule simplifies the process of coupling color develop- 
ment. Since the synthetic polymer is the sole emulsion binder in a 
given emulsion layer, and since the polymer contains an abundance of 
color coupler nuclei as part of its chemical structure, high efficiency 
in the process of color coupling is achieved. A practical consequence 
of achieving high efficiency in dye generation is an enhanced compact- 
ness of dye-image deposit as defined by the silver image itself. Defini- 
tion and sharpness of image in three-color prints is apparent as a re- 



The successful utilization of the principle of dual-purpose emulsion 
binders has resulted in the development of a new motion picture color 
positive stock designated Du Pont Type 275. As one would expect, 
the evolution of a photographic color process based entirely on the 
complete replacement of gelatin with synthetic polymers has involved 
a variety of complex research problems. 

At the outset, it was necessary to undertake the chemical synthesis 
of polymeric materials having properties permitting their use in place 
of gelatin, the traditional photographic emulsion medium. Up to this 
time, no completely satisfactory non-gelatin materials had been de- 
veloped, even for application in the black-and-white field. In ad- 
dition to properties permitting their use as emulsion media, the further 
requirement was imposed upon the new polymers that they must 
function in the capacity of couplers for use in a process of color pho- 
tography. Three different color-forming binders were in fact re- 
quired, each capable of producing a different subtractive color com- 
ponent, yet having related qualifications for the other important role. 

The complex chemistry of these new polymers, while a broad sub- 
ject in itself, is being touched upon only briefly in this paper because 
of limitations of space. For those to whom this subject is of interest, 
further treatment is to be given elsewhere.* 

Having produced synthetic color-forming binders, a second essen- 
tial step leading to the construction of a color film product was the 
development of methods for making the emulsions and for coating the 
new materials. In a number of significant respects, the technology 
which emerged differs markedly from practices which have become 
conventional for gelatin systems. 

The new synthetic color-forming binders, in general, satisfy the 
exacting requirements for the medium in which the photographic silver 
halide is suspended. These requirements have been generalized by 
Mees 3 as follows: "It must keep the emulsion grains perfectly dis- 
persed to eliminate clumping and consequent granularity of the 
photographic image; it must be stable for a long period of time, so 
that both the undeveloped and the processed emulsion are reasonably 
permanent; it must impart no undesirable photographic characteris- 
tics to the emulsion grains ; it must be such that it can be handled in a 
relatively simple and yet accurately reproducible manner, so that 
emulsions can be made and coated by practicable procedures; and, 

* This portion of the work has been carried out mainly in the Experimental Sta- 
tion Laboratories of the Chemical Dept., Wilmington, Del. 


finally, it must allow the penetration of processing solutions without 
impairment to its strength, toughness, and permanence after the 
processing operations are completed." 

While a number of synthetic binders, including polyvinyl alcohol, 
have been studied as replacements for gelatin, those which have 
proven practical for color emulsions are acetals of polyvinyl alcohol. 
Commercial methods have been developed for the manufacture of 
those selected for use. The coupling and other groups introduced by 
acetalization modify the properties of the parent polyvinyl alcohol to 
the extent that practical binders result. A typical structural unit for a 
synthetic color-forming polyvinyl acetal binder can be represented : 

J \ ? /V \ V /f \ 

-C-CH 2 -C-CH 2 -C-CH 2 -G-CH 2 -C-CH 2 -C-CH 2 -C-CH 2 - 
OH /, OH 6 \OH /, Q H J> \OH /, 

X v i s "i ^ i ^ '"L 




It is to be kept in mind that any coupling nucleus and developing 
agent that gives a dye of suitable color characteristics can be chosen, 
since mobility of the couplers and solubility of the final dye is not a 
problem. Also, while the coupling groups are very close to the site of 
the development reactions, the rate of reactivity is influenced by the 
coupling nucleus; so that this factor remains important in their selec- 
tion. This latter is particularly important in multilayer structures 
where lower layers are less accessible to processing solutions than the 
upper layers. 

Although the new technique quite nicely avoids some of the prob- 
lems previously involved in coupling color development, it introduces 
new problems in the manufacture of photographic products. Gelatin, 
although having certain disadvantages, has some properties very use- 
ful to the photographic emulsion maker. These are its ability to form 
thermally reversible gels at convenient temperatures, and its contribu- 
tion to the establishment of suitable photographic properties. So far, 
it has been difficult to synthesize binders that combine these proper- 
ties so that, if the step were desirable for some other reason, the syn- 
thetic binder could be substituted directly for gelatin. Nevertheless, 
it has been possible to devise a technology for making and handling 
photographic emulsions and films utilizing the polyvinyl acetal color- 
forming binders. 


In the making of a photographic emulsion the first step establishes 
the physical characteristics of the silver halide grains. This involves 
precipitation of silver halides in the presence of the binder, or part of 
it, and changing the size distribution of the grains by an operation in 
which the larger grains grow at the expense of the smaller ones. 

It is next necessary to wash the emulsion to remove soluble reaction 
products of the first step which cannot be tolerated in the finished film. 
The reaction involved in the precipitation can be very generally 
stated as being silver nitrate plus alkali metal halide yielding silver 
halide and alkali metal nitrate: 

AgNO 3 + MX -> AgX + MN0 3 

where the alkali metal nitrate (MNOs) represents the reaction product 
which must be removed. In gelatin technology it is the custom to 
chill the emulsion so that it forms a firm gel and in this state reduce it 
to small pieces by physical means, thereby increasing the amount of 
surface. The small pieces can then be leached with cold water to re- 
move the undesirable reaction products which are extremely soluble in 
comparison to the silver halides. Since the new synthetic color-form- 
ing binders do not readily form gels in the manner of gelatin, it has 
been found convenient to devise chemical procedures for precipitating 
the emulsions and washing them. 

In addition to being a necessary step in the handling of gelatin 
emulsions, the maintenance of the emulsions in the chilled state is 
necessary to slow down the rate of decomposition of the gelatin, which 
is a material so susceptible to putrefaction under some conditions that 
preservatives in addition to the chilling are required to make the pro- 
cedures practicable. The new synthetic color-forming binders are by 
their chemical nature free of this difficulty. 

After removal of the undesired by-products, the emulsion is ready 
for its final treatments before coating. With gelatin emulsions the 
washed pieces are melted by the application of heat before proceeding 
with the finishing treatments. With the new synthetics, the washed 
emulsion is simply redissolved. The actual finishing operations them- 
selves, including extra-sensitizing procedures, addition of sensitizing 
dyes where needed, adjustment of pH, addition of wetting agents, etc., 
are strictly comparable for the two systems. 

In the coating and drying operations necessary for applying the 
finished emulsions to film base, the differences between the two sys- 
tems are again very apparent. With gelatin emulsions, the applied 
layer is chilled immediately after coating to form a firm gel and 
gradually dried from this gel state. The solidification procedure is 


necessary to prevent flowing of the thin layers with formation of an 
uneven and nonuniform coating. With the new synthetic color-form- 
ing binders, chemical procedures have been devised for coagulating 
the emulsion to a nonfluid state before drying begins. This has been 
accomplished by treating the base suitably at the time it is made. 


During the work on development of the synthetic color-forming 
binders and their application in Du Pont Type 275, it became appar- 
ent that the new polymers could confer upon a photographic product a 
number of unusual physical characteristics, not all of which were 
visualized at the outset. Some of these properties may have im- 
portant technical significance in the future, and they will be reviewed 
before discussing the properties and performance of the stock itself. 

Brittleness and lack of flexibility, particularly at low humidities, are 
well-known properties of gelatin. In contrast, the synthetic polymers 
which have been discussed exhibit a high degree of flexibility, tough- 
ness and resistance to abrasion. A composite polymer monopack struc- 
ture, for example, composed only of functional photographic layers, 
each a few ten thousandths of an inch thick and having a total thick- 
ness of only 0.001 in. can be shown to be strong and self-supporting 
after solvent removal of the base support. 

In an experiment such as that just described, and in analogous treat- 
ments with a variety of solvents, it is apparent that the developed dyes 
as well as the color-forming binders are polymeric and accordingly 
completely insoluble. Thus, problems in connection with waxing, 
cleaning or polishing operations will be minimized. The tendency of 
gelatin to respond rapidly to thermal changes is apparent to a much 
lesser degree in the case of the synthetic polymers. While the implica- 
tions of this property have not been fully explored as yet, one ad- 
vantage that may accrue is processing at elevated temperatures. 


Du Pont Type 275 is a monopack subtractive color film having all 
dye-generating layers superposed on one side of the support and re- 
quiring no step wise processing or transfer operations. It utilizes cyan 
(minus-red), magenta (minus-green), and yellow (minus-blue) syn- 
thetic color-forming binders of the type discussed above. The struc- 
ture is shown in Fig. 1. The functions of the various layers on the 
base are as follows : 

1. The top emulsion layer, unsensitized to other than blue light, 
receives the magenta image from the green analysis record by printing 




with blue light. The yellow dye that is present and which distributes 
itself throughout the film absorbs the blue light as it passes through 
and prevents it from exposing the bottom layers which, since they 
contain silver halides, are also blue-sensitive. The yellow dye dis- 
solves out during processing. 

2. The separator layers prevent interlay er effects, not those usually 
caused by migration of the coupler molecules, since in this film these 
have been immobilized by making them an integral part of the binder, 
but those caused by migration of oxidized developer molecules be- 
tween adjacent layers. 

3. The middle emulsion layer, sensitized only to red light in ad- 
dition to its inherent blue sensitivity, receives the cyan image from the 
red analysis record by printing with red light. 









Fig. 1. Structure of Du Pont Type 275 Color Film, 
Release Positive. 











4. The bottom emulsion layer, sensitized only to green light as well 
as retaining its inherent blue sensitivity, receives the yellow image 
from the blue analysis record by printing with green light. 

5. The substratum layers anchor the emulsion and backing layers to 
the film base. 

6. The antihalation backing coated on the rear of the film absorbs 
any light passing through the emulsions into the base, so that it can- 
not be reflected from the back surface of the film and cause halation. 
During development the backing dye is decolorized, and later on, dur- 
ing the washing steps, the entire backing layer dissolves off spon- 
taneously without mechanical scrubbing. The spectral absorption of 
the backing is shown in Fig. 2, there being ample density at all wave- 
lengths where protection from halation is required. 

It is to be noted here that the layers other than the emulsion layers 




are also prepared using synthetic polymers, thus completely eliminat- 
ing gelatin from a commercial photographic product. The physical 
properties, such as water sensitivity and swelling, of all the different 
polymers used have been balanced in order to make a film of satisfac- 
tory characteristics. At the same time, the permeability of the layers 





550 600 650 700 


Fig. 2. Spectral absorption of the antihalation backing. 

TYPE 275 

Y C M 





Fig. 3. Color reproduction with Du Pont Type 275. 

B =Blue 
N = Neutral 

C = Cyan 
R = Red 

G = Green 
W = White 

M = Magenta 
Y = Yellow 

to processing solutions has been maintained at a high level in order to 
keep the lag between the start of development in the outermost layer 
and the lower layers at a minimum. Fortunately, the physical proper- 
ties of the synthetic color-forming binders can be balanced by adjust- 
ment of the number of color-forming nuclei present and by the intro- 
duction of other groups. 




Fig. 4. Wedge spectrogram showing sensitivity peaks of magenta, yellow and cyan 



Since this film is designed for printing from color separation records, 
it is possible to have a layer arrangement and sensitivity as shown 
without regard to the kind of light originally required to make the 

various records. This is explained 
in Fig. 3. The arrangement chosen 
has been adopted in the interest 
of resolution (sharpness) with the 
various images positioned in order 
of importance, namely, the green- 
record image in magenta on top, 
the cyan (red-record) image next 
and the yellow (blue-record) last. 
A wedge spectrogram showing 
the spectral response of the com- 
plete film is shown in Fig. 4. This 
shows peaks at 440, 550, and 710 
mju (millimicrons) for the ma- 
genta, yellow and cyan layers, 
respectively. The peak at 440 mju 
in the magenta layer is produced 

1.0 20 


Fig. 5. Sensitometric curves for indi- 
vidual layers and neutral scale. 

by the instrument used, the peak for the silver halide used actually 
being at about 390 mju. 

Photographic characteristics have been adjusted to permit printing 
from color analysis records having equal effective contrasts. The 
sensitometric curves of the various layers for a standard set of de- 
veloping conditions using the developing agent p-aminodiethylaniline 
are shown in Fig. 5, with densities measured at the wavelength of 
maximum absorption of each dye. A similar set of curves was 
selected as a goal of the work by calculations made from a neutral 
sensitometric curve of desired characteristics (gamma = 2.5, straight- 
line densities to 2.8) and the absorption characteristics of the dyes from 
the color-forming binders selected for use. The absorption curves of 




the various dyes and their contributions to a neutral density of 1.0 
are given in Fig. 6. 


500 6OO 



Fig. 6. Spectral characteristics of component dyes and resulting neutral. 


To make pictures on Du Pont Type 275 Color Film, Release Positive, 
three-color separation negatives must be printed onto the one release 
film in such manner that the positive image of each color separation is 
recorded only in the proper dye-generating layer. Naturally the 
three dye images in each frame must be superimposed in register. Be- 
cause of the complex nature of the stock, only limited variation of con- 
trast can be achieved in processing, so the primary control of contrast 
and balance lies in adjusting the contrast of the negatives. These re- 
quirements mean, for the present, that Type 275 must be printed in a 
register printer from three black-and-white separation negatives, each 
printing exposure being made through a narrow-pass filter so that the 
image will be confined to the appropriate subtractive color. 

Any black-and-white three-color separation negatives may be used 
to print Type 275 provided they have the following characteristics : 

1. Proper color separation ; 3. Good register; 

2. Correct orientation for 4. Acceptable sharpness; 

same-side printing; 

5. Reasonably fine grain; 

6. Appropriate contrast. 


Negatives meeting these specifications may be derived from such 
existing or proposed methods as: stripping film, beam-splitting 
cameras, filter-wheel cameras, and separations from monopack color 

To amplify these characteristics somewhat, the first means that 
each separation must record only the intended color aspects of a scene. 
The red negative, for example, must not respond to blue or green in 
addition to red, otherwise color degradation will result. Most di- 
rectly exposed negatives meet this requirement quite well. Separa- 
tions from color positives are often acceptable, but may be improved 
by masking. 

Correct orientation for one-side printing requires the mirror-image 
reversal of at least one of the negatives obtained with beam-splitting 
cameras or stripping film. Since this can be done in optical printers of 
the type commonly used in the industry, this is usually no prob- 

The requirements of good register, acceptable sharpness, and fine 
grain are common to all color work. The excellent resolving power of 
Type 275 emphasizes the need for good register and fine grain, be- 
cause there is almost no diffusion of the image to cover up poor register 
or coarse grain in the negatives. 

Figure 7 illustrates the contrast requirements for negatives to be 
used with Type 275. It shows a comparison between a black-and- 
white fine-grain release positive and the gray-scale or "equivalent" 
density characteristic curve of the color film. A black-and-white 
print with full tonal range may encompass the densities 0.15 to 2.3, 
which correspond to a density range of 1.1 in the negative. In the 
case of color films, experience shows that the density range in the print 
is typically somewhat greater, perhaps 0.15 to 2.8. The density scale 
in the negative must be correspondingly somewhat higher, about 1.45, 
which is 1.3 times as great. Thus, while negatives for black-and- 
white use typically have a 7 of 0.65 to 0.7, negatives for contact print- 
ing of Type 275 should be at 7 0.85 to 0.90. This factor 1.3 is almost 
exactly the gain in contrast of projection printing compared with con- 
tact printing. Thus, negatives of the same contrast as normally used 
in black-and-white practice may be printed optically on Type 275. 

Inasmuch as the contrast of the new color film is subject to only 
relatively small adjustment via processing, the gamma of the negatives 
is the major variable by which the contrast of the final image may be 
controlled. If negative gamma is not appropriate to begin with, it 
will be necessary to alter it by duping. 







1.0 2.0 




Fig. 7. Comparison of H&D curves for color and black-and-white positives, 
illustrating contrast requirements for negatives. 

Exposing Filters 

Having obtained a suitable set of separation negatives, the next 
step is to expose each record into the proper emulsion layer of the 
print film. Figure 1 has shown the relation between the spectral 
sensitivity of each layer and the dye generated in that layer by color 
development. The top layer is blue-sensitive and forms the magenta 
dye, thus it must be printed from the green-record negative. The 
middle layer is sensitive to red light and its image is cyan ; so it is ex- 
posed from the red-record negative. The bottom layer develops 
yellow; so it receives the blue-record image by printing with green 
light, to which the layer is sensitive. The color sensitivity of each 
layer bears no required relationship to the color of the subtractive dye 
it carries, for it is simply a means of confining each exposure to the 
proper layer. The proper exposing color is obtained by using a nar- 
row-pass filter over the light source. 

The wedge spectrogram (Fig. 4) shows the spectral region to which 
each layer is sensitive, aiding the selection of exposing filters. Sensi- 
tivity of the top, magenta-forming emulsion extends from the ultra- 




violet to about 490 mju. The response of the green-sensitive emulsion 
begins at about 495 rmz, so, clearly, the blue filter must cut off at 
wavelengths shorter than this value. The green sensitivity extends 
to about 590 mju, with a peak at 550 m/*. The red sensitivity begins at 
600 m/i and peaks at 710 m/z. The selection of practical filters in- 
volves finding ones with maximum transmissions of the desired colors 
and minimum leak of undesired wavelengths. The most efficient set is : 

Blue Corning 5113, half -thickness; 

Green Defender 60G; 

Red Corning 2403, full-thickness. 

There is enough variation among filter batches so that individual 
filters should be checked photographically, or on a spectrophotometer. 
Similarly, filters in production use should be checked from time to 
time for constancy. 

580 600 


Fig. 8. Spectral curve of safelight filter. 

Safelight Filters 

Related to color sensitivity is the question of safelights. Re-examin- 
ing Fig. 4 suggests that the most efficient safelight would transmit 
freely at one or other of the two gaps in sensitivity, at 495 mju or at 
600 m/z. The 600-ni;u gap has been chosen because it is a little 
wider. A spectrophotometric curve of the safelight filter designed for 
this film is shown in Fig. 8. An infrared-absorbing component is de- 
sirable in the safelight, because the cyan layer has considerable infra- 
red sensitivity to wavelengths which most organic dyes begin to 
transmit rather freely. With the filter illustrated, a five-minute ex- 


posure is safe with an illumination of 0.02 ft-c. While this is not quite 
as bright as safelights commonly used with black-and-white positive, 
it is very bright indeed for a film having essentially panchromatic 

It is interesting to note in passing that a monochromatic source 
emitting the wavelength of one of the gaps in sensitivity would make a 
very efficient safelight indeed. Although the "D" lines from a 
sodium arc do not fall quite at an ideal wavelength, they do produce 
very high illumination for a given* level of "safety." 

Printing Equipment and Illumination Requirements 

A registration printer, either optical or contact, is required for ex- 
posing the picture images onto Type 275, because this operation in- 
volves three successive printings from separation negatives. Any 
conventional printer of this class may be used; the only additional 
specification over black-and-white printing is that filters must be 
placed between the light source and film. 

Any standard light-change device may be used. These include 
traveling mattes, apertures, and lamp voltage control. It should be 
noted in the latter connection that perfect freedom is permissible in 
varying lamp voltage. Since the exposures are made through narrow- 
pass filters, this process does not require that the source be operated at 
a fixed color temperature. Whatever the mechanism, it is desirable to 
have fine printer-point steps to give maximum control over color bal- 

Regarding illumination requirements in the printer: a 500-w in- 
candescent lamp 10 in. from the film plane, with spherical mirror, in a 
contact printer with Jfo~ sec exposure time gives ample exposure. 


Processing of Du Pont Type 275 Color Film consists of four chemi- 
cal treatments, with water washes or rinses between. The steps are 
shown in Table I. 

TABLE I. Processing Steps for Du Pont Type 275 Color Film. 

1. Develop . . 
2. Wash .... 
3. First fix . . 

.... 10-12 min 
.... 1-2 min 
.... 6 min 

6. Wash 
7. Second fix . . . . 
8. Wash 

. . 4 min 
. . 4 min 
. . 10 min 

4. Wash .... 
5. Bleach* . . . 

.... 5 min 
.... 5 min 

9. Dry 

Note: If sulfiding of track is employed, this may be done after a 1-min wash 
following Step No. 5. Processing then proceeds to Step No. 6. 


After the film has been exposed in the printer, it is developed in a 
color developer. Here a silver image is developed in each layer, and 
concurrently the final dye images are also generated. Following a 
wash, the film passes to the first fixer, where all silver halide not used 
in the primary image is dissolved. The next treatment is a bleach 
which converts the silver image to silver ferrocyanide, which is dis- 
solved by the second fixer. If the sound track is to be sulfided, this 
may be done following the bleach, before the second fix. Water 
washes are necessary between chemical treatments to avoid excessive 
contamination of one solution with another, which might lead to 
shortened solution life and the possibility of stain on the film. As 
with other films, a final wash is used to remove all processing chemicals 
from the emulsions. Drying is accomplished in the usual manner. 

The processing times given in Table I are based on 70 F solution 
temperatures. While the temperature of the developer should be held 
quite constant ( =*=}/ deg) for uniform results, temperature control is 
not particularly critical in the subsequent steps, since the reaction in 
each treatment is to be carried to completion. 

Temperatures other than 70 F may be used if more convenient. 
Higher temperatures lead to shorter processing times, and may be 
very desirable in cases where machine capacity is limited. Type 275 
has been processed successfully with all solutions at 90 F or above, 
without the need for special hardening treatments. 

The times listed in Table I for processing steps beginning with the 
first wash have been selected as the minimum, in the interest of de- 
veloping-machine compactness. Additional treatment time is per- 
missible if machine capacity is available, and would provide a wider 
safety factor to assure complete reaction. 

Processing Solutions 

Table II gives formulas of solutions for processing. 

The developer formula given in Table II should be considered 
approximate, and may vary for individual processing machines, de- 
pending upon conditions of agitation, etc. The reason for this is 
evident from a consideration of the complex structure of the film. It is 
obvious that the three emulsion layers do not have the same accessi- 
bility to developer, image formation naturally progressing more 
rapidly in the upper layers, which the developer reaches first. Thus, 
proper contrast balance to give a gray-scale is achieved only under a 
restricted range of processing conditions. Over-all contrast cannot be 
altered significantly by a simple change of development time as in 
black-and-white film, because a departure from the proper developing 




time, with no other compensating change, throws the contrast rela- 
tions of the three layers into incorrect balance. Since different de- 
veloping machines do not give identical results, some adjustments 
may be needed to obtain a balanced development. 

In general, the composition of processing solutions, particularly the 
developer, must be maintained to closer tolerances than are allowable 
in black-and-white photography. It is imperative that replenishment 
be based on accurate analytical techniques. It is recommended that 
the developer be replenished continuously, though the other solutions, 
which have wider tolerances, may receive batch wise additions of re- 

TABLE II. Formulas for Processing Type 275 Color Film. 


Monohydrochloride . . 2.5 g 
Sodium Sulfite, anhyd. . 10. Og 
Sodium Carbonate, mono- 

hyd 47. Og 

Potassium Bromide ... 2 . g 
Water to make 1.01 

pH = 10.5 (approx.) 


Potassium Ferricyanide . . 100 g 

Boric Acid 10 g 

Borax 5 g 

Water to make 11 

pH = 7.0 to 7.5 

First Fixer 

Hypo, crystals 240 g 

Sodium Sulfite, anhyd. . . 15 g 

Borax 18 g 

Acetic Acid, 28% 43 cc 

Potassium Alum 20 g 

Water to make 11 

pH =4.5 (approx.) 

For use: dilute 1 part solution 

to 2 parts water 

Second Fixer 

Hypo, crystals 200 g 

Water to make 11 

pH = 8.0 (approx.) 

plenisher. Analytical techniques are available, but their description is 
outside the scope of this paper. 

Processing Machine Design 

Figure 9 illustrates a form of continuous developing machine for 
Type 275 Color Film. The sketch is schematic and is intended to sug- 
gest only the proper tank arrangement. As far as details of design 
and construction are concerned, this particular process does not neces- 
sitate any features different from good black-and-white practice. In 
fact, a black-and-white machine may be converted provided it has 
enough tanks to allow proper arrangement of solutions. 





Sound reproduction problems peculiar to multilayer color films with 
dye-image tracks have been the subject of several papers in the 
JOURNAL. Special techniques have had to be developed for dye- 
image sound tracks, for it is now generally recognized that they can- 
not be used directly with the red-sensitive phototubes which are now 
standard for theater motion picture projectors. 4 It is the universal 
characteristic of organic dyes suitable for three-color images that they 
are quite transparent in the near-infrared spectral region to which the 
868 phototube has its greatest response. The result of trying to use 
such a combination inevitably is weak modulation and poor signal-to- 
noise ratio. 










Fig. 9. Schematic arrangement of processing machine. 

Two solutions to this dilemma have been found. One is to use a 
magenta-dye sound track in conjunction with a phototube with an 
S-4 photosurface. 5 - 6 The spectral response of such a phototube when 
illuminated with a tungsten lamp has its peak at about 530 m/*, 
which is near the maximum absorption of the magenta dye. Thus, 
variations in magenta density modulate the phototube strongly. A 
second solution is to convert the silver image formed in the sound 
track during development into silver sulfide, which is relatively non- 
transparent in the near infrared, hence can be used with standard red- 
sensitive phototubes. Both of these methods can be utilized success- 
fully with Du Pont Type 275 Color Film. The former has the ad- 
vantage of simplicity, for it requires no extra treating steps; it also 
has certain technical superiorities. The latter has the advantage of 
expediency; it yields tracks which may be played with the present, 
existing phototubes. 


Magenta Sound Tracks 

When a phototube with S-4 sensitivity is to be used for reproduction, 
a sound track is applied to Type 275 Color Film in a manner com- 
pletely analogous to black-and-white practice. The sound negative is 
simply printed on the color positive using the blue-exposing filter to 
confine the image to the magenta-forming emulsion, and the film is 
developed as described without any additional treatments to the 
sound tr&ck. 

The top density and contrast of the magenta image, while dictated 
by picture requirements, are quite appropriate for sound reproduction. 
In particular, the high resolving power of the magenta emulsion con- 
fers good high-frequency response to a magenta track. 

Because the contrast of the magenta image as "seen" by a blue- 
sensitive phototube is somewhat lower than black-and-white release 
positive, variable-density negatives intended for printing on Type 275 
should have somewhat higher contrast than is used with black-and- 
white positive. A Du Pont Type 228 negative developed to a IIB 
control gamma of about 1.2 will yield a magenta track with minimum 
intermodulation for positive track densities in the neighborhood of 0.6 
density. Likewise, variable-area sound negatives should have higher 
track density than if intended for black-and-white use. In cross- 
modulation tests a magenta track, printed from a Du Pont Type 201 
negative developed to a IIB gamma of 3.5 and having a track density 
of 2.5, had optimum cancellation for a 1.15 density. 

Actual intermodulation data for magenta sound tracks are repre- 
sented by the solid curve of Fig. 10, and cross-modulation data appear 
in Fig. 11. These distortion measurements were made with a 1P37 
phototube in the sound reproducer, and the track densities read with a 
blue-sensitive phototube in the densitometer. 

Sulfided Sound Tracks 

A sulfided sound track is produced on Type 275 by an edge treat- 
ment following the bleach, but preceding the second fix. At this point 
in the processing the original silver image has been converted to silver 
ferrocyanide by the bleach. This compound reacts very rapidly with 
sodium sulfide to form silver sulfide, which has the desired opacity to 
near-infrared radiation. 

The picture image also contains silver ferrocyanide, so it is obvious 
that the entire film should not be treated with sodium sulfide. There- 
fore, an applicator which treats only the sound track with the sulfiding 
solution must be used. Such applicators are not novel, and many 


forms have been used successfully. To keep the solution from diffus- 
ing to undesired areas, its viscosity is raised by the addition of a 
thickening agent. 

Sulfiding Solution 

Distilled water ( 125 F) 750 cc 

Sodium Carboxymethylcellulose, medium viscosity 20 g 

Sodium Sulfide, nonahydrate 63 g 

Water to make , i 1 

Stir thoroughly with a mechanical stirrer and filter while hot. Cool to room 
temperature before using. 

The film should receive a 30- to 60-sec water rinse following the 
bleach bath to eliminate excess ferricyanide solution. At this time, 
the film is removed from the machine and prepared for the sound- 
track beading operation. An air blow-off should be used to remove 
excess liquid from the surface of the film. Best results are obtained 
when the emulsion is partly dried by passing the film through a drying 

The film travel is so arranged that a developing time of one full 
minute is allowed following the application of the sulfiding solution. 
At the end of this period, the sound track area is subjected to a small, 
high-velocity water stream directed to wash the treating solution 
toward the perforations. This removes the excess sulfiding solution. 
The film is now returned to the machine for completion of the normal 
process, the next treating bath being the second fix. 

Sound tracks to be sulfided are exposed in a slightly different way 
than are magenta tracks. While it would be desirable from the point 
of view of sharpness to print the track in the top layer only, the 
amount of silver in the magenta emulsion alone is too low to produce a 
silver sulfide track of the desired density. Thus, it is necessary to 
utilize the lower emulsions by printing with white light, even though 
some loss of high-frequency response results. 

The following operating conditions were found at the Du Pont 
laboratories to give satisfactory results. A variable-area negative re- 
corded on Du Pont Type 201 Sound Recording film was exposed to 
give a negative track density of 2.5 with the film processed to gamma 
3.5. This track was printed onto Type 275 with unfiltered incandes- 
cent light and sulfided as described. Cancellation of 30 db or more 
occurred at positive track densities in the neighborhood of 1.2. A 
variable-density sound negative recorded on Du Pont Type 226 
processed to a IIB gamma of 1.5 yielded a color sound print with 
minimum intermodulation at positive track densities about 0.6. 









--SULFIDE TRACK TYPE 226 NEG. H B V = 1.50, DENS. = 0.5 


0.3 9-4 0.5 0.6 0.7 0.8 

Fig. 10. Intermodulation curves for magenta 
and sulfide variable-density sound tracks. 




^ -20 











Fig. 11. Cross-modulation curves for magenta 
and sulfide variable-area sound tracks. 

The dashed curves in Figs. 10 and 11 show actual intermodulation 
and cross-modulation data, respectively, for sulfided sound tracks 
played with a standard 868 phototube. The track densities likewise 
were measured with a red-sensitive phototube as the receiving element 
of the densitometer. 


Volume output of both magenta and sulfided sound tracks is nor- 
mal, being within 1 or 2 db of a standard silver track. Signal-to- 
noise ratio is somewhat better than black-and-white positive, attribut- 
able to the exceedingly fine grain structure of the colored image. 
Following are typical signal-to-noise ratios comparing the three types 
of track for variable-area recording without noise reduction: 

Fine-grain Release Positive (black-and-white) 36 db; 

Type 275 magenta track (IP 37 cell) 38 db; 

Type 275 sulfide track (868 cell) 40 db. 

High-frequency reproduction with color sound tracks is somewhat 
inferior to silver tracks. This is particularly true of sulfided tracks. 
The loss is caused partly by the high negative track density require- 
ment and partly by the fact that a sulfided track utilizes all three 
of the emulsion layers of the color film. At9000 cycles a magenta 
variable-area track is about 2 db from a silver track, while the sul- 
fided track is about 7 db from the silver reference. Some high- 
frequency boost during recording may be necessary with the latter 


The developments reported in this paper were made possible 
through the combined efforts of many Du Pont research workers. In 
addition to the work in the laboratories of the Technical Division, 
Photo Products Dept., Parlin, N.J., extensive contributions have been 
made by the Chemical Department, Experimental Station, Wilming- 
ton, Del. 


1. R. Fischer, German Patent No. 253,335, 1912. 

2. R. Fischer and H. Siegrist, "The formation of dyestuffs by means of oxidation 

with irradiated silver halides," Phot. Korr., vol. 51, pp. 18-21, 1914; pp. 208- 
211, 1914. 

3. C. E. K. Mees, The Theory of the Photographic Process, p. 59, Macmillan, New 

York, 1942. 

4. R. Gorisch and P. Gorlich, "Reproduction of color film sound records," Jour. 

SMPE, vol. 43, pp. 206-213, September 1944. 

5. A. M. Glover and A. R. Moore, "A phototube for dye image sound track," 

Jour. SMPE, vol. 46, pp. 379-386, May 1946. 

6. R. D. Drew and S. W. Johnson, "Preliminary sound recording tests with 

variable-area sound tracks," Jour. SMPE, vol. 46, pp. 387-404, May 1946. 

An Improved Video System 
For Television Studios 


SUMMARY: A new video system and a new arrangement of television 
studios have been devised at WOR-TV. This system adds considerably 
to the flexibility of a television station by permitting combination of any 
of the station's cameras in any combination in several studios for program- 
ming. This is accomplished by using a separate "camera control center" 
where all camera control operations are carried out. Individual studio 
control rooms provide for program direction and switching of the cameras 
for a particular program. 

IN DESIGNING a television studio system at the present time the 
primary consideration is to make the whole system as flexible as 
possible first, because television broadcasting is still a relatively 
new art, and therefore constantly changing as new ideas are developed ; 
and second, because program requirements for television shows vary 
constantly from day to day, each program making different demands 
on the technical facilities. 

Purposing to make a television studio as adaptable as possible in 
all its phases, the Television Engineering Department of WOR pro- 
ceeded with the design of its new television studios, located on West 
67th Street in New York. The arrangement of the studios in this 
building incorporate several novel ideas, of which we shall discuss 
here the separation of the camera control operators from the program 
direction and switching center. 

The camera control operators for all studios are centrally located in 
one room called the Camera Control Center. Thus the program 
director is not distracted by having to look over the shoulders of 
the technical operating personnel or by any confusion arising from 
their being in the same room. 

In this system the program director has directly in front of him, at 
his console, monitors on each of his local cameras, plus two preview 
monitors for switching up the cameras from some other source which 
might be cut in to his program. In addition, the switching control 
panel is located on this desk for the video switcher who sits beside the 
program director. 

By centrally locating all camera control units in one place, further 

PRESENTED: October 16, 1950, at the SMPTE Convention in Lake Placid, N. Y. 




Fig. 1A. Floor plan of first floor showing studios and control rooms; (1) Con- 
trol Room A, (2) Announce Booth D, (3) Announce Booth E, (4) Control Room 
C, and (5) Control Room B. 


B A L.C N Y 











'CTB.U 5 


34 * 16 

1 6^ MASTER. 
V - 1 51X24 





Fig. IB. Floor plan of second floor showing projection and master control rooms 
and upper part of studios. 




advantages from a technical point of view are realized. First, main- 
tenance operations on the equipment are no longer hampered during 
rehearsal and program periods by the presence of program personnel 
in the same room, and, in case of trouble, replacement of equipment 
in the Camera Control Center is greatly facilitated. Secondly, the 
flexibility of operation is greatly increased. For example, any com- 
bination of the eight studio cameras and three film cameras can be 

o o o 




C L I E. 



Fig. 2 A. Typical floor plan of studio control looms at 67th Street. 

Video Console Components 
preview monitor 1 (7) preview monitor 2 

camera monitor 1 
camera monitor 2 
camera monitor 3 
camera monitor 4 
line monitor 

(8) director's intercom panel 

(9) projection room remote control 

(10) technical director's intercom 

(11) switching panel 

(12) receiver monitor 

used in any combination on any program switched through any of the 
program control rooms. This flexibility is further increased by the 
use of a camera cable patch panel in the Camera Control Center, 
enabling any of the studio camera controls to be patched to cables 
leading to any of the main studios or announce studios. Thirdly, 
centralized camera control eliminates the electrical delay problem 
which arises when several studios are located at different distances 
from the Master Control Room. 




With this new arrangement, it is possible to realize a saving in the 
number of operating personnel assigned to camera control func- 

The control room space in the 67th Street Studios is divided as 
shown in Figure 1A. Three studio or program control rooms are pro- 
vided here, each of which has identical facilities. One of these control 
rooms is normally used with each of the two large studios. The con- 
trol room floor level is about two feet above the studio floor. A large 
window in each control room permits good viewing of the studio. 

Fig. 2B. Director's console in studio contiol room. 

The third control room, Studio "C" Control Room, is normally used 
for handling of remote or film programs. Film inserts on remotes are 
easily handled in this control room by routing the remote signal 
through the Studio "C" switching system. In addition, all station 
breaks and film spot announcements are handled in the Studio "C" 
Control Room. A typical studio control room arrangement is shown 
in Figs. 2A and 2B. Here the program director's console with its 
switching control panel is located on the left with the audio control 
console and turntables on the right. The camera switching is con- 


trolled from each director's console by means of a push-button panel 
containing sixty momentary contact push buttons. These control 
circuits operate video switching relays located centrally in the Camera 
Control Center for all studios. 

Each of the three studio camera switching systems consists of a re- 
lay bank of twelve inputs with five outgoing channels. This permits 
the handling of up to eight local camera signals and three remote com- 
posite signals through any studio control room. In addition, the 
twelfth input to the switching system is used as an "Effects" input for 
switching in a super-position, lap dissolve or other type of effect, as 
required. The five outgoing channels feed the two preview monitors 
in the director's console, the effects mixer amplifier, and the main pro- 
gram output of that studio. 

The space above the program control rooms on the second floor 
contains the Film Projection Room and the combination Camera 
Control Center and Master Control Room (Fig. IB) . In the Camera 
Control Center all of the camera control units are located together in a 
large U-shaped console, In addition, forty equipment racks house the 
associated amplifiers, power supplies, synchronizing generators and 
the power supplies for all the studios. 

In the camera control section, eight studio camera controls, each 
with its picture monitor and oscilloscope and two line monitors form 
the section facing the studios. A special feature of the Camera Con- 
trol Center is the camera cable patch panel shown on the right side in 
Fig. 3. This is mounted on the wall directly adjacent to the camera 
control units themselves. The sockets mounted on these panels cor- 
respond to cables leading to the various studios. The camera cable 
pigtails that plug into these sockets correspond to the eight studio 
camera control units. Thus, the eight camera controls can be dis- 
tributed in any combination among the fifteen circuits to the various 
studios depending upon program requirements. This adds greatly to 
the flexibility of the over-all system and makes it possible to take care 
of almost any special requirement. Furthermore it reduces the total 
number of camera chains required in such an aggregate of studios. 

It is thus that, in case of trouble with the equipment during a pro- 
gram or rehearsal, it is very easy to patch in a spare camera control 
unit so that the equipment giving trouble may be released for main- 

In addition to the patching of the camera control units to any of 
the studios on the camera cable patch panel, it is, of course, necessary 
to patch the video outputs of the camera controls on the coaxial jack 
panels which are mounted in racks adjacent to the program control 




room where the switching is done. Also, of course, it is important 
that tally light circuit information and intercommunication facilities 
for the particular camera and camera control follow the proper pro- 
gram console. For this purpose a special tally and intercommunica- 
tion patch panel is provided in the equipment racks directly above the 
coaxial jack panels. Thus, any camera control can be set up on any of 
the three camera switching systems for tally and intercommunication 
control. In this way a camera control operator may plug in a headset 
to any one of the control sections which he is working and have com- 
plete two-way intercommunication with the video switcher down in 

Fig. 3. Camera Control Center with camera cable patch panel at right. 

the program control room and also with the cameramen. In addition, 
on a separate earphone he may listen to the program audio from that 

The film camera control sections are located adjacent to and on the 
left of the studio camera control sections, forming part of the over-all 
console. Switching of these and patching is handled in a manner 
similar to that of the studio camera controls with the exception that 
the camera cable from each camera control is tied directly to its cor- 
responding film camera in the projection room. 

The film projection room is next door to the Camera Control 
Center. Here three iconoscope film camera chains are installed, each 




with a mirror multiplexer system to combine optically several sources 
of film or slides on one camera. Two 35-mm film projectors, two 16- 
mm film projectors, a Gray "Telop" (opaque projector), three 2X2 
slide projectors and a straight opaque projector comprise the film 
room projection equipment. Picture monitors for each of the camera 
chains are located in equipment racks beside the projectors, enabling 
the projectionist to line up each of his film cameras before being 
switched on the air. An intercommunication system associated with 
each film camera enables the projectionist to be in communication 
with anyone of the three program control rooms to which he is as- 

Fig. 4. Master control switching console. 

In the Camera Control Center is also located the master control 
switching equipment. The master control switching section is the 
left wing of the large U-shaped console (Fig. 4). This equipment is all 
relay-operated and comprises a switching system for handling six 
studios to four outgoing channels. It is a preset switching system for 
both audio and video and is arranged for either simultaneous or inde- 
pendent audio-video switching as required. Provision is made for 
either tripping all four outgoing channels with a single master switch 
or any group of them together. A picture monitor and oscilloscope 


as well as an audio monitor and audio level meter are assigned to each 
outgoing channel so that levels on each outgoing channel can be set 
independently as well as monitored. 

Two synchronizing signal generators are provided, one for standby 
use in case of trouble. An RCA Genlock unit has also been installed 
which permits the line-by-line as well as field-by-field phasing of the 
local synchronizing signal generator with an incoming remote signal. 
This permits the use of a remote signal in lap dissolving and super- 
position with local cameras and it has been found to be particularly 
effective with film commercials inserted during remote programs. 

The new WOR-TV Studios have been in use since about the first of 
February. During this time it has been found that the facilities have 
met almost all of the requirements put upon them by the Program 
Department. Flexibility in meeting these requirements has been ap- 
parent in producing programs that would have been very difficult 
under more usual conditions. Tn addition, centralized camera control 
operation has resulted in helpful economies in both manpower and 

Infrared Photography 
With Electric-Flash 



SUMMARY: Electric-flash sources produce infrared as well as visible and 
ultraviolet light. Using only the infrared portion, the guide number for in- 
frared film is equal to or higher than the guide number for Kodachrome using 
the visible portion from the same source. Factors affecting the infrared out- 
put are discussed. Described is a small airplane instrument panel photo- 
graphic recorder using an infrared filter over the flashtube to avoid distrac- 
tion of the pilot. The short-duration infrared flash gives very readable 
records on 16-mm film at one-second intervals. 

THE EXISTENCE of the infrared portion of the spectrum has been 
known since the early part of the last century, and infrared 
photographic records have been produced for over sixty years. 
However, it was not until 1931, when advances in sensitive materials 
were made, that the application of infrared photography became 
practical. 1 Photography using wavelengths longer than 13,000 
A (Angstrom units) is still difficult. Present-day applications de- 
pend on: (1) the ability of infrared radiation to penetrate haze; 
(2) differential absorption and reflectance of these long wavelengths 
by different materials; and (3) the inability of the human eye to 
respond to infrared light. These properties have led to the use of 
infrared in aerial photography, camouflage detection, crime detec- 
tion, medicine, botany, and many other fields. 

For some subjects the source of infrared radiation is the subject it- 
self, but in all other cases some external source must be used. Com- 
mon infrared sources are sunlight, incandescent light, arc lights and 
photoflash lamps. Less common, although not new, is the use of 
electric-flash (stroboscopic or high-speed lights) which makes possible 
the practice of infrared photography using very short exposures. It 
is this type of infrared light source and its application with which 
this paper is concerned. 

Twelve or fifteen years ago electric-flash techniques were little 
known, but today their use is commonplace. 2 Several papers on this 
subject have been published in this Journal. 3 - 4 However, it is less 
generally known that an electric-flash source emits a great deal of 
infrared radiation. Figure 1, a typical spectral distribution curve of 
PRESENTED: April 26, 1950, at the SMPTE Convention in Chicago. 





a commercial electric-flash tube, shows that the energy reaches a 
maximum in the blue, decreases to a minimum in the near-infrared, 
and then increases again. The energy per 100-A band in the infrared 
is approximately half the energy per 100-A band in the visible. The 
total energy in the visible region of 4,000 to 7,000 A is only about 
three times the total energy in the infrared between 7,200 and 8,700 A, 
but this figure will vary considerably between tube types and with 
other circuit constants. Thus it is seen that electric-flash is an effec- 
tive source of infrared light. 

5000 7000 9000 10000 


Fig. 1. Spectral energy distribution 
for a typical flashtube; data supplied 
by General Electric Co. 

SA o oo o w Z 




IN 88A 


















Fig. 2. Characteristics of the red-sensitive 
photocell and Wratten 88A filter used for 
measuring infrared radiation. The shaded 
region represents the response of the com- 



The spectral energy distribution of Fig. 1 is for a standard com- 
mercially available electric-flash tube used for normal black-and- 
white or color photography. The question immediately arises as to 
the possibility of changing the tube design or operating conditions to 
increase the infrared output. Factors which should be investigated 
are voltage, capacity, tube loading, tube dimensions, gas pressure 


and types of gases. To obtain some indication of the effect of voltage 
and capacity on infrared output, three types of tubes were investi- 
gated. Infrared light measurements were made with a special 
integrating-type light-meter 5 designed for use with electric-flash, 
which employed a red-sensitive photocell and a Wratten 88A infrared 
filter. 6 Figure 2 shows the sensitivity curve of the photocell and the 
transmission curve of the 88A filter. This filter has a transmission 
of less than 0.1% below 7,200 A. The overlapping of the two curves 
(shown by the shaded area) represents the region of response of this 
combination of filter and photocell. This region extends from 7,200 
to 12,000 A. 

Using the methods of measurement described, three different types 
of flashtubes were investigated to determine the effect of voltage and, 
to a limited extent, energy loading on the infrared output. The 
General Electric Co. FT-110 is a new flashtube designed for 1,000-v 
operation in portable electric-flash equipment. Its infrared effi- 

2 in 3 


LU tr 2 


FT-214' ~"T~-- 


Fig. 3. Effect of operating voltage on the 

infrared efficiency of several flashtubes. 

500 1000 1500 2000 

ciency, as shown in Fig. 3, is nearly three times as high at 500 as at 
2,000 v when operated at 50 w-sec (watt-seconds), and approximately 
twice as high when operated at 12.5 w-sec. The CAA (Civil Aero- 
nautics Administration) flashtube shows an infrared efficiency more 
than twice as high at 500 as at 2,000 v. This tube is a very small 
quartz lamp designed specifically for an infrared instrument recorder 
described later in this paper. The GE FT-214 is a standard 2,000-v 
flashtube used for portable and semiportable flash equipment. Its 
infrared efficiency remains approximately constant over this voltage 
range. From these limited data, it is probably safe to state that for 
maximum infrared efficiency a flashtube should be designed for as 
low voltage as is consistent with proper starting characteristics and 
flash duration. 

The other design factors of tube dimensions, gas pressure and type 
of gas also may very well affect the infrared efficiency and should be 
investigated as they have been for the visible region. 2 




Although the energy in the infrared approaches the energy in the 
visible region of an electric-flash source, the lower sensitivity of in- 
frared film makes the over-all combination of electric-flash source, 
filter and infrared film considerably slower than the electric-flash 
panchromatic film combination. However, it is as fast as, or slightly 
faster than, the electric-flash Kodachrome combination. These 
speeds can be considered in terms of the commonly used guide num- 

Fig. 4. Electric-flash infrared photograph; note "freezing" of fan blade. 

bers (aperture or /-stop multiplied by distance). For example, a 
small electric-flash unit might have a guide number of 150 to 200 for 
super-speed panchromatic film, a guide number of 30 for Kodachrome, 
and a guide number of 30 to 50 for infrared film when a Wratten 
88 A filter is used. It is understood that if a Wratten 87 filter, which 
cuts off at approximately 7,600 A, is used, the guide number will be 
reduced. It can be said in general that the infrared guide number 
will be at least equal to the Kodachrome guide number for a given 
electric-flash unit. Any subject that can be photographed in color 
can probably be photographed in infrared with identical equipment. 


A commercial 10,000-w-sec flash unit having a guide number for 
Kodachrome of 250 could be used to photograph in infrared an area 
of several thousand square feet at an aperture of //3.5. A press 
photographer's portable strobe unit could be used for figure length 
photographs at perhaps //2.5, and finally, working toward smaller 
and smaller areas, infrared photomicrographs with electric-flash 
should be well within the realm of possibility. Figure 4 is an in- 
frared photograph taken with a flash unit using a guide number 
nearly 50% higher than the guide number which normally would be 
used for Kodachrome. 


An ideal application of electric-flash as a source of infrared light 
is represented in a small automatic data recorder or "cockpit ob- 
server" for photographically recording instrument readings in air- 
craft during flight tests. The development of an instrument for 
use on extended flights was sponsored by the Civil Aeronautics 
Administration and participated in by the Fairchild Camera Co. 
through a contract with the CAA in 1936. It was extended through 
a contract with Eastman Kodak in 1941. The intention was to 
filter out all visible light and photograph with infrared in order 
to remove all possibilities of distracting the pilots. Incandescent 
sources were originally used, but the relatively long exposures re- 
quired resulted in blurred images due to aircraft vibration. Elec- 
tric-flash as the infrared light source offered the possibility of eliminat- 
ing this defect. 

A CAA Contract with Edgerton, Germeshausen & Grier, Inc., in 
1943 resulted in two special 16-mm cameras and an experimental 
110-v a-c electric-flash unit. The cameras constructed by the East- 
man Kodak Co. were adaptions of the standard 16-mm Magazine 
Cine Kodak. The spring motors were removed and electric motor- 
drives substituted. Fast-acting overriding shutters were incor- 
porated to give a short exposure despite the slow operating speed of 
two frames per second. This was necessary to minimize the effect of 
daylight which would have superimposed an additional exposure of 
long duration. Contact synchronizers with zero time delay were also 
incorporated for synchronizing the electric-flash. The cameras 
proved to be very satisfactory for this application. 

In 1947 the application requirements of the equipment were changed 
to those of a flight test recorder, in particular, for small aircraft. 
Inasmuch as the principal requirement was to have a source of 


illumination essentially invisible to the pilot, C. W. Wyckoff of the 
Edgerton, Germeshausen & Grier staff undertook to re-evaluate the 
relative merits of working in the ultraviolet or infrared regions of 
the spectrum. It was determined that although the over-all effi- 
ciency in the ultraviolet region was considerably greater than that 
in the infrared, it would have been necessary for the pilot to wear 
light yellow glasses to cut out the blue portion of the spectrum. In 
addition to this disadvantage, the contrast using ultraviolet light was 
considerably less than that obtained with infrared light. Further, 
the ultraviolet light would cause the luminescent dial paints to glow, 

Fig. 5. Quartz flashtube for infrared recorder compared in size 
to a 35-mm film cartridge. 

which might or might not have been a disadvantage. Having con- 
cluded that an infrared source had the greater advantage, the CAA 
outlined certain requirements for the recorder : (1) area to be covered, 
11 X 14 in.; (2) camera to instrument panel distance, 36 in.; 
(3) maximum aperture, //1. 9; (4) picture rate, one per second; (5) 
power source, 12-v battery; and (6) minimum size and weight. 

A few simple photographic tests with standard flashtubes indi- 
cated that a special flashtube would be desirable. Two seemingly 
incompatible factors enter into the design of a flashtube which is to 
be operated at high repetitive rates. If it is small, it will overheat 
due to the average power input, but will have high efficiency. If it is 




increased in size to handle the average power, it will have low effi- 
ciency, requiring in turn a higher average power input. When these 
conditions apply, it is advantageous to construct the tube of quartz 
which can be operated at a much higher temperature level than glass, 
and a compromise must be made between efficiency and power 
handling ability. For the CAA recorder, a small quartz U-shaped 
tube (see Fig. 5) was designed which had a reasonable efficiency and 
was able to handle the necessary power without overheating. Its 
size permitted the use of a very small reflector without sacrificing re- 
flector efficiency. As shown in Fig. 3, its infrared efficiency is about 
50% higher at 500 than at 1,000 v and hence a voltage operating level 
of 475 to 500 v was chosen. This voltage level has the advantage 
of simplifying the power supply design for an airborne application. 
The dimensions of the tube and gas pressure are such as to give re- 
liable starting at this voltage. A special mounting base protects the 
fragile graded quartz-to-tungsten seals and fits into a standard 

Fig. 6. Schematic diagram of 
infrared electric-flash instrument 

fluorescent starter socket which locks the flashtube in place, but at 
the same time makes it easily removable. Adequate spacing for the 
spark lead is obtained by placing its termination well up on the side 
of the base. Using a Wratten 88 A filter over a small reflector, 4 in. 
in diameter, a Wratten 88 filter. over the lens, and infrared film, an 
input to the flashtube of 12 to 14 w-sec gives the illumination required 
for the recorder. 


The circuit (Fig. 6) has several features of particular interest. 
Inasmuch as the primary power source was to be an aircraft 12-v 
battery, one of three types of conversion was available: (1) vibrator, 
transformer and rectifiers; (2) d-c to a-c inverter, transformer and 
rectifiers; or (3) 12-v d-c to 450-v d-c dynamotor. Dynamotors are 
generally preferred to vibrators in aircraft and eliminate the need for 
transformers and rectifiers and therefore were chosen as the means 




of voltage conversion. This would not have been practicable if the 
voltage requirement had been much above 500 v. 

Two 220-juf 475-v electrolytic capacitors (C-l and C-2) are con- 
servatively operated in series to give an energy storage of about 12 w- 
sec. Chokes CH-1 and CH-2 both serve to prevent "hold-over" in 
the flashtube FL-1. A "hold-over" in a flashtube is a continuous 
glow which occurs if an attempt is made to recharge the capacitors 
too rapidly. The tube is in a highly ionized state after being flashed 

Fig. 7. Complete infrared electric-flash instrument recorder, showing power 
unit, camera and lamphouse assembly, cables and a mock-up of instruments. 

and, if sufficient current is supplied, the tube will not deionize but will 
remain conducting and act as a short circuit across the power supply. 
A choke (CH-1) acts as a high impedance immediately after flashing 
to limit the current and thus prevent "hold-over." On the other 
hand, over a period of one second it acts as a relatively low im- 
pedance and therefore allows the capacitors C 1 and C-2 to charge 
fully before the next flash. An additional choke (CH-2) in series 
with the flashtube also tends to prevent "hold-over" by causing a 


slight reversal of voltage, thereby allowing the tube to deionize. The 
amount of reversal must be limited to a small value to prevent dam- 
age to the electrolytic discharge capacitors. 

To prevent inverse output voltage from the dynamotor, which in 
turn would cause failure of the electrolytic capacitors, it is necessary 
to prevent operation of the dynamotor in the event that the polarity 
of the battery voltage is ever reversed. This is accomplished by 
placing a selenium rectifier in series with the coil of the main control 
relay RY-1 which prevents its closing if the polarity of the voltage 
is incorrect. 

The cameras described previously have had new 12-v d-c governor- 
controlled motor-drives installed to operate them at one frame per 
second to within an accuracy of about 1%. The nonreversal pro- 
tection described above also prevents the camera motors from driving 
the film in the wrong direction. 

Mechanical Design 

Figure 7 shows the complete unit (total weight 18 Ib), which in- 
cludes the power supply, camera, lamphouse assembly and cables. 
A test bank of instruments is also shown. The motor-drive, governor, 
and lamphouse are attached to the camera, but the lamphouse may 
be removed for side-lighting by releasing two small snaplocks. 
Slack cable coiled inside the motor-drive housing allows operation of 
the lamp up to 30 in. from the camera. The total weight of the 
camera unit (including camera, motor-drive, and lamphouse as- 
sembly) is approximately 4 Ib. The camera and lamphouse are 
shown in Fig. 8 which illustrates how a Wratten 88 A filter is held in 
place over the reflector by a grooved rubber ring that also serves to 
prevent any unfiltered light leaking around the edge of the filter. 
It will be noted that the camera also has an infrared filter (a Wratten 
88) and that the flashtube is located so that the legs of the U are in a 
horizontal plane. This orientation tends to spread the light in the 
horizontal direction corresponding to the longer axis of the 16-mm 
frame. The power supply is a rectangular aluminum box with all 
components attached to the cover for easy servicing. The dyna- 
motor is placed on top for maximum cooling. 

In the laboratory the unit has operated very satisfactorily. Figure 
9 is an infrared photograph of a test instrument bank taken with the 
recorder at 36 in., an angle of 30 deg, and an aperture of //1. 9. As 
data can be taken readily from exposures made at //2.8, it should 
be possible to photograph areas slightly larger than specified, and by 
the use of two recorders, more extensive instrument panels can be 




Fig. 8. Camera and lamphouse assembly showing infrared filter 
and rubber retaining ring removed. 

Fig. 9. Enlargement of section of 16-mm film made with the CAA 
infrared electric-flash instrument recorder. 


observed. At the operating rate of one per second a series of over 
2,000 photographs covering a flight test of over 30 min is possible 
without reloading. 

In contrast to special recording systems which require a separate 
bank of test instruments, this device can be readily installed in the 
cockpit of even the smallest aircraft to record data from the standard 
instrument panel. However, without the infrared niters it is also 
adaptable for use in recording much larger aircraft instrument in- 
stallations located where the flashing light would not distract the 
pilot. Installation and testing of this infrared recorder in a DC-3 
airplane is now being carried on by the CAA. 


Commercial electric-flash techniques can be used to provide in- 
frared light for infrared photography. Guide numbers at least as 
large as those for Kodachrome can easily be obtained. The efficiency 
of electric-flash as an infrared source is generally highest for a tube 
designed to operate at the lowest practicable voltage. An aircraft 
instrument recorder, or "cockpit observer," employing electric- 
flash as an infrared source has been designed and shows satisfactory 
performance in laboratory tests. 


1 . W. Clark, Photography by Infrared, 2d ed. , John Wiley & Sons, New York, 

1946. (An extensive treatment of infrared photography.) 

2. H. E. Edgerton, ' 'Photographic use of electrical discharge flashtubes," J. 

Opt. Soc. Amer., vol. 36, no. 7, pp. 390-399, July 1946. 

3. H. E. Edgerton, "Electrical-flash photography," Jour. SMPE, vol. 52, pp. 

8-23, Mar. 1949. 

4. K. J. Germeshausen, "New high-speed stroboscope for high-speed motion 

pictures," Jour. SMPE, vol. 52, pp. 24-34, Mar. 1949. 

5. H. E. Edgerton, "Light-meter used with electronic flash," /. Photo. Soc. 

Amer., Part II, Photographic Science and Technique, pp. 6-10, January 

6. W ratten Light Filters, p. 69, Eastman Kodak Company, Rochester, N.Y., 


Magnetic Sound Film 
Developments in Great Britain 

BY 0. K. KOLB 


SUMMARY: The introduction of magnetic sound film recording and repro- 
ducing apparatus into Great Britain is described as well as the types and 
characteristics of magnetic film available. Details and the general circuit 
arrangement of the apparatus are given together with a description of special 
apparatus which has been used for adding a visible signal indication record 
to the invisible magnetic sound track. Apparatus which has been evolved 
for the bulk wiping of magnetic film stock is also described as well as ex- 
perience gained with different types of magnetic film joints. 

MOST PEOPLE are familiar with the Telegrafone, the name which 
Poulsen gave to his magnetic sound recorder and reproducer, 
in which a steel wire was run past a magnetic head to which signals 
were fed and thereby recorded on the wire, the signals being subse- 
quently picked up again from the steel wire by means of the same or a 
similar magnetic head. 

The steel wire recording system developed very slowly, mostly due 
to the absence of any large developments in the electronic field, but 
when, after World War I, thermionic valve amplifiers, good micro- 
phones and good loudspeakers were developed, the steel wire was again 
taken up but only for speech and signal (e.g., Morse) recording. It 
did not find its way into the motion picture industry chiefly for two 

Firstly, the quality was not very good and could not compare 
with that of photographic sound. 

Secondly, the speed of the steel wire was much too fast to allow it 
to be synchronized satisfactorily with the picture. 

The Magnetophone: Successor to the Telegrafone 

Some years before World War II, Pfleumer in Germany had 
developed the use of magnetic iron oxide powders on tapes as a mag- 
netic sound recording medium and the Allgemeine Elektricitats 
Gesellschaft had marketed a machine for the use of it; but even this 
new apparatus, called the Magnetophone 1 did not open the field of 
application to the motion picture industry, since the quality was still 
of the standard of a dictating machine. 

A CONTRIBUTION: Submitted March 9, 1950 



High-Frequency Bias for Magnetophone Tape 

However, further developments during World War II, especially in 
Germany by Braunmuhl and Weber, 2 who applied the high-frequency- 
bias method of recording to the oxide tape, changed the situation 
materially, as they produced a magnetic sound record greatly im- 
proved and far superior to the old d-c 3 or even a-c 4 biased Telegrafone 
and the Magnetophone, d-c operated until that time. It was now 
possible to obtain a quality comparable and even superior to any other 
recording means with a frequency response up to 20,000 cycles/sec 
and with the tape running at the reasonable speed of about 30 in./sec. 

Application to the Motion Picture Industry 

The motion picture industry now started to take an interest in these 
new developments in magnetic recording and the possibilities of intro- 
ducing magnetic sound into the industry were examined in different 

In Great Brit